Allied Telesis Dust Collector AT S63 User Manual

Management  
Software  
AT-S63  
Features Guide  
AT-S63 Version 2.2.0 for the AT-9400 Layer 2+ Switches  
AT-S63 Version 3.0.0 for the AT-9400 Basic Layer 3 Switches  
613-000801 Rev. A  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Contents  
Preface ............................................................................................................................................................ 17  
Chapter 1: Overview ...................................................................................................................................... 29  
3
Download from Www.Somanuals.com. All Manuals Search And Download.  
Contents  
Chapter 2: Enhanced Stacking .....................................................................................................................55  
Chapter 3: SNMPv1 and SNMPv2c ...............................................................................................................65  
Chapter 4: MAC Address Table ....................................................................................................................71  
Chapter 5: Static Port Trunks .......................................................................................................................75  
Chapter 6: LACP Port Trunks .......................................................................................................................81  
Chapter 7: Port Mirror ....................................................................................................................................93  
Chapter 8: File System ..................................................................................................................................99  
Chapter 9: Event Logs and the Syslog Client ...........................................................................................105  
4
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Chapter 10: Classifiers ................................................................................................................................ 109  
Chapter 11: Access Control Lists .............................................................................................................. 119  
Chapter 12: Class of Service ...................................................................................................................... 131  
Chapter 13: Quality of Service ................................................................................................................... 139  
Chapter 14: Denial of Service Defenses .................................................................................................... 161  
5
Download from Www.Somanuals.com. All Manuals Search And Download.  
Contents  
Chapter 15: IGMP Snooping .......................................................................................................................175  
Chapter 16: MLD Snooping .........................................................................................................................179  
Chapter 17: RRP Snooping .........................................................................................................................183  
Chapter 18: Ethernet Protection Switching Ring Snooping ....................................................................187  
Chapter 19: SNMPv3 ....................................................................................................................................197  
Chapter 20: Spanning Tree and Rapid Spanning Tree Protocols ...........................................................213  
6
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Chapter 21: Multiple Spanning Tree Protocol ........................................................................................... 225  
Chapter 22: Port-based and Tagged VLANs ............................................................................................. 247  
Chapter 23: GARP VLAN Registration Protocol ....................................................................................... 261  
Chapter 24: Multiple VLAN Modes ............................................................................................................. 273  
Chapter 25: Protected Ports VLANs .......................................................................................................... 279  
7
Download from Www.Somanuals.com. All Manuals Search And Download.  
Contents  
Chapter 26: MAC Address-based VLANs ..................................................................................................285  
Chapter 27: Internet Protocol Version 4 Packet Routing .........................................................................299  
Chapter 28: BOOTP Relay Agent ................................................................................................................331  
Chapter 29: Virtual Router Redundancy Protocol ....................................................................................337  
8
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Chapter 30: MAC Address-based Port Security ....................................................................................... 349  
Chapter 31: 802.1x Port-based Network Access Control ........................................................................ 355  
Chapter 32: Web Server .............................................................................................................................. 381  
Chapter 33: Encryption Keys ..................................................................................................................... 387  
9
Download from Www.Somanuals.com. All Manuals Search And Download.  
Contents  
Chapter 34: PKI Certificates and SSL ........................................................................................................397  
Chapter 35: Secure Shell (SSH) ..................................................................................................................413  
Chapter 36: TACACS+ and RADIUS Protocols .........................................................................................423  
Chapter 37: Management Access Control List ..........................................................................................431  
Appendix A: AT-S63 Management Software Default Settings .................................................................439  
10  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Appendix B: SNMPv3 Configuration Examples ........................................................................................ 475  
Appendix C: Features and Standards ....................................................................................................... 481  
11  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Contents  
Appendix D: MIB Objects ............................................................................................................................489  
Index ..............................................................................................................................................................509  
12  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Figures  
13  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Figures  
14  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Tables  
15  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Preface  
This guide describes the features of the AT-9400 Layer 2+ and Basic  
Layer 3 Gigabit Ethernet Switches and the AT-S63 Management Software.  
This preface contains the following sections:  
Caution  
The software described in this documentation contains certain  
cryptographic functionality and its export is restricted by U.S. law. As  
of this writing, it has been submitted for review as a “retail encryption  
item” in accordance with the Export Administration Regulations, 15  
C.F.R. Part 730-772, promulgated by the U.S. Department of  
Commerce, and conditionally may be exported in accordance with  
the pertinent terms of License Exception ENC (described in 15  
C.F.R. Part 740.17). In no case may it be exported to Cuba, Iran,  
Iraq, Libya, North Korea, Sudan, or Syria. If you wish to transfer this  
software outside the United States or Canada, please contact your  
local Allied Telesis sales representative for current information on  
this product’s export status.  
17  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
ˆ Section V: Spanning Tree Protocols  
ˆ Section VI: Virtual LANs  
ˆ Section VII: Routing  
ˆ Section VIII: Port Security  
ˆ Section IX: Management Security  
ˆ Appendices  
19  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Preface  
Product Documentation  
For overview information on the features of the AT-9400 Switch and the  
AT-S63 Management Software, refer to:  
ˆ AT-S63 Management Software Features Guide  
(PN 613-000801)  
For instructions on starting a local or remote management session, refer  
to:  
ˆ Starting an AT-S63 Management Session Guide  
(PN 613-000817)  
For instructions on installing or managing stand-alone switches, refer to:  
ˆ AT-9400 Gigabit Ethernet Switch Installation Guide  
(PN 613-000357)  
ˆ AT-S63 Management Software Menus Interface User’s Guide  
(PN 613-50570-00)  
ˆ AT-S63 Management Software Command Line Interface User’s Guide  
(PN 613-50571-00)  
ˆ AT-S63 Management Software Web Browser Interface User’s Guide  
(PN 613-50592-00)  
For instructions on installing or managing a stack of AT-9400 Basic Layer  
3 Switches and the AT-StackXG Stacking Module, refer to:  
ˆ AT-9400 Stack Installation Guide  
(PN 613-000796)  
ˆ AT-S63 Stack Command Line Interface User’s Guide  
(PN 613-000777)  
20  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Where to Go First  
Allied Telesis recommends that you read Chapter 1, “Overview” on page  
29 in this guide before you begin to manage the switch for the first time.  
There you will find a variety of basic information about the unit and the  
management software, like the two levels of manager access levels and  
the different types of management sessions.  
This guide is also your resource for background information on the  
features of the switch. You can refer here for the relevant concepts and  
guidelines when you configure a feature for the first time.  
21  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Preface  
Starting a Management Session  
For instructions on how to start a local or remote management session on  
the AT-9400 Switch, refer to the Starting an AT-S63 Management Session  
Guide.  
22  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Document Conventions  
This document uses the following conventions:  
Note  
Notes provide additional information.  
Caution  
Cautions inform you that performing or omitting a specific action may  
result in equipment damage or loss of data.  
Warning  
Warnings inform you that performing or omitting a specific action  
may result in bodily injury.  
23  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Preface  
Where to Find Web-based Guides  
The installation and user guides for all Allied Telesis products are  
available in portable document format (PDF) on our web site at  
www.alliedtelesis.com. You can view the documents online or download  
them onto a local workstation or server.  
24  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Contacting Allied Telesis  
This section provides Allied Telesis contact information for technical  
support as well as sales and corporate information.  
Online Support You can request technical support online by accessing the Allied Telesis  
Knowledge Base: http://kb.alliedteleisn.com. You can use the  
Knowledge Base to submit questions to our technical support staff and  
review answers to previously asked questions.  
Email and For Technical Support via email or telephone, refer to the Support &  
Services section of the Allied Telesis web site: www.alliedtelesis.com.  
Telephone  
Support  
Returning Products for return or repair must first be assigned a return materials  
authorization (RMA) number. A product sent to Allied Telesis without an  
RMA number will be returned to the sender at the sender’s expense.  
Products  
To obtain an RMA number, contact Allied Telesis Technical Support  
through our web site: www.alliedtelesis.com.  
Sales or You can contact Allied Telesis for sales or corporate information through  
our web site: www.alliedtelesis.com. To find the contact information for  
your country, select Contact Us -> Worldwide Contacts.  
Corporate  
Information  
Management New releases of management software for our managed products are  
available from the following Internet sites:  
Software Updates  
ˆ Allied Telesis web site: www.alliedtelesis.com  
ˆ Allied Telesis FTP server: ftp://ftp.alliedtelesis.com  
FTP client software is required to download new software from the Allied  
Telesis FTP server using your workstation’s command prompt.  
Furthermore, you must log in to the server. The user name is “anonymous”  
and the password is your email address.  
25  
Download from Www.Somanuals.com. All Manuals Search And Download.  
             
Preface  
26  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Section I  
Basic Operations  
The chapters in this section contain background information on basic  
switch features. The chapters include:  
Section I: Basic Operations  
27  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
28  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 1: Overview  
Layer 2+ and Basic Layer 3 Switches  
The switches in the AT-9400 Gigabit Ethernet Series are divided into two  
groups:  
ˆ Layer 2+ Switches  
– AT-9408LC/SP  
– AT-9424T/GB  
– AT-9424T/SP  
ˆ Basic Layer 3 Switches  
– AT-9424T  
– AT-9424Ts  
– AT-9424Ts/XP  
– AT-9448T/SP  
– AT-9448Ts/XP  
The switches of the two groups offer many of the same features and  
capabilities. However, there are a couple of significant differences. For  
instance, the Internet Protocol Version 4 packet routing feature is only  
supported on the Basic Layer 3 switches and is the reason for the group’s  
name. For a list of the supported features, refer to Table 1. The switches  
are numbered in the table as follows:  
Layer 2+ switches:  
ˆ 1 - AT-9408LC/SP  
ˆ 2 - AT-9424T/GB  
ˆ 3 - AT-9424T/SP  
Basic Layer 3 switches:  
ˆ 4 - AT-9424T  
ˆ 5 - AT-9424Ts  
ˆ 6 - AT-9424Ts/XP  
ˆ 7 - AT-9448T/SP  
ˆ 8 - AT-9448Ts/XP  
The Stack column lists the features supported in a stack of Basic Layer 3  
switches and the AT-StackXG Stacking Module. For more information,  
(Y = supported feature)  
30  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Table 1. AT-9400 Switch Features  
Layer 2+ Switches  
(Version 2.2.0)  
Basic Layer 3 Switches  
(Version 3.0.0)  
Stack1  
1
2
3
4
5
6
7
8
-
Basic Operations  
Local management  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Remote Telnet  
management  
Remote Secure Shell  
management  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Remote web browser  
management  
TCP/IP pings  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Enhanced stacking  
Simple Network Time  
Protocol (SNTP)  
SNMPv1 and SNMPv2  
Port statistics  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Static port trunks  
Link Aggregation Control  
Protocol (LACP) trunks  
Port mirroring  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Advanced Operations  
Y2  
File system  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y3  
Y
Event logs  
TFTP client  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Syslog client  
Y
Classifiers  
Access control lists  
Class of Service  
Y
31  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 1: Overview  
Table 1. AT-9400 Switch Features  
Layer 2+ Switches  
(Version 2.2.0)  
Basic Layer 3 Switches  
(Version 3.0.0)  
Stack1  
-
1
2
3
4
5
6
7
8
Quality of Service  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Denial of service defenses  
Snooping Protocols  
Internet Group  
Y
Y
Y
Y
Y
Y
Y
Y
Management Protocol  
(IGMP) snooping  
Multicast Listener  
Discovery (MLD) snooping  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Router Redundancy  
Protocol (RRP) snooping  
Ethernet Protection  
Switching Ring (EPSR)  
snooping  
SNMPv3  
SNMPv3  
Y
Y
Y
Y
Y
Y
Y
Y
Spanning Tree Protocols  
Spanning Tree Protocol  
(STP)  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Rapid Spanning Tree  
Protocol (RSTP)  
Multiple Spanning Tree  
Protocol (MSTP)  
Virtual LANs  
Port-based and tagged  
VLANs  
Y
Y
Y
Y
Y
Y
Y
Y
Y
32  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Table 1. AT-9400 Switch Features  
Layer 2+ Switches  
(Version 2.2.0)  
Basic Layer 3 Switches  
(Version 3.0.0)  
Stack1  
1
2
3
4
5
6
7
8
-
802.1Q-compliant and  
non-802.1Q-compliant  
multiple VLAN modes  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
GARP VLAN Registration  
Protocol  
Y
Y
Y
Y
Y
Y
Y
Y
Protected ports VLANs  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
MAC address-based  
VLANs  
Internet Protocol Routing  
Internet Protocol version 4  
packet routing  
Y
Y
Y
Y
Y
One routing interface4  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Virtual Router Redundancy  
Protocol  
BOOTP and DHCP clients  
BOOTP relay agent  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Port Security  
MAC address-based port  
security  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
802.1x port-based network  
access control  
Management Security  
Encryption keys  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Public Key Infrastructure  
(PKI) certificates and  
Secure Sockets Layer  
(SSL) protocol  
33  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 1: Overview  
Table 1. AT-9400 Switch Features  
Layer 2+ Switches  
(Version 2.2.0)  
Basic Layer 3 Switches  
(Version 3.0.0)  
Stack1  
-
1
2
3
4
5
6
7
8
Remote Secure Shell  
management  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
TACACS+ and RADIUS  
authentication  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Management access  
control list  
1. Basic Layer 3 switches using version 3.0.0 of the management software and the AT-StackXG Stacking  
Module.  
2. The only accessible file system in a stack is on the master switch.  
3. The master switch has the only active event logs in a stack.  
4. Used to assign the switch or stack an IP configuration.  
34  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
AT-S63 Management Software  
The AT-9400 Switch is managed with the AT-S63 Management Software.  
The software comes preinstalled on the unit with default settings for all the  
operating parameters of the switch. If the default settings are adequate for  
your network, you can use the switch as an unmanaged unit.  
Note  
The default settings are listed in Appendix A, “AT-S63 Management  
You can access the management software on the switch several different  
ways. You can manage the switch locally (out-of-band) using the Terminal  
Port on the front panel or over a network (in-band) using a Telnet or  
Secure Shell client, or a web browser. For further information, refer to  
The management software has three management interfaces -- a menus  
interface, a command line interface, and a web browser interface. You can  
use any of the interfaces to perform basic configuration procedures. But  
some of the newer and more complex features, such as Virtual Router  
Redundancy Protocol (VRRP), must be configured with the command line  
interface. For more information, refer to “Management Interfaces and  
There are two current versions of the management software, Version 2.2.0  
and Version 3.0.0. Version 2.2.0 is for the Layer 2+ switches:  
ˆ AT-9408LC/SP  
ˆ AT-9424T/GB  
ˆ AT-9424T/SP  
Version 3.0.0 is for the Basic Layer 3 switches:  
ˆ AT-9424T  
ˆ AT-9424Ts  
ˆ AT-9424Ts/XP  
ˆ AT-9448T/SP  
ˆ AT-9448Ts/XP  
Note  
Do not install version 3.0.0 on a Layer 2+ switch.  
35  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 1: Overview  
Management Interfaces and Features  
The AT-S63 Management Software has three management interfaces:  
ˆ Menus interface  
ˆ Command line interface  
ˆ Web browser interface  
You can use the menus and command line interfaces from a local  
management session through the Terminal Port on the switch or remotely  
with a Telnet or Secure Shell client. The web browser interface is used  
from remote HTTP and HTTPS sessions using a web browser.  
You can configure all the features and parameters of the switch from the  
command line interface. However, the menus interface and the web  
browser interface are limited in the number of functions that you can  
perform. For example, you can configure the basic port settings (e.g.,  
speed and duplex mode) for any interface, but VRRP can only be  
configured using the command line interface.  
Table 2 lists the functions supported by the three management interfaces.  
Note  
A stack of AT-9400 Basic Layer 3 Switches and the AT-StackXG  
Stacking Module must be configured from the command line  
interface.  
(Y = supported feature)  
Table 2. Management Interfaces and Features  
Command Line  
Interface  
Web Browser  
Interface  
Menus Interface  
Basic Operations  
Switch’s name, location, and  
contact  
Y
Y
Y
Manager and operator passwords  
Date and time (manual and SNTP)  
Rebooting a switch  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
TCP/IP pings  
36  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Table 2. Management Interfaces and Features  
Command Line  
Interface  
Web Browser  
Interface  
Menus Interface  
Enhanced stacking  
Y
Y
Y
Y
Y
Y
SNMPv1 and SNMPv2  
community strings  
Port parameters  
Port statistics  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
MAC address table  
Static MAC addresses  
Static port trunks  
Link Aggregation Control Protocol  
(LACP) trunks  
Port mirroring  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Baud rate of the Terminal Port  
Management console timer  
Telnet server  
Console startup mode  
Advanced Operations  
Y1  
Y2  
File system and configuration files  
Y
Y
Format flash memory  
Y
Y
File uploads and downloads  
Y
Y
Y3  
Y
Y
Y
Y
Y
Y
Event logs  
Y
Syslog client  
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Classifiers  
Access control lists  
Class of Service  
Quality of Service  
Denial of service defenses  
37  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 1: Overview  
Table 2. Management Interfaces and Features  
Command Line  
Menus Interface  
Interface  
Web Browser  
Interface  
Snooping Protocols  
Internet Group Management  
Protocol (IGMP) snooping  
Y
Y
Y
Y
Y
Y
Y
Y
Multicast Listener Discovery  
(MLD) snooping  
Router Redundancy Protocol  
(RRP) snooping  
Ethernet Protection Switching  
Ring (EPSR) snooping  
SNMPv3  
SNMPv3  
Y
Y
Y
Spanning Tree Protocols  
Spanning Tree Protocol (STP)  
Y
Y
Y
Y
Y
Y
Rapid Spanning Tree Protocol  
(RSTP)  
Multiple Spanning Tree Protocol  
(MSTP)  
Y
Y
Y
Virtual LANs  
Port-based and tagged VLANs  
Y
Y
Y
Y
Y
Y
802.1Q-compliant and  
non-802.1Q-compliant multiple  
VLAN modes  
GARP VLAN Registration Protocol  
Protected ports VLANs  
Y
Y
Y
Y
Y
Y
Y
MAC address-based VLANs  
38  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Table 2. Management Interfaces and Features  
Command Line  
Interface  
Web Browser  
Interface  
Menus Interface  
Internet Protocol Routing  
Routing interfaces  
Y
Y
Y
Y
Y
Static routes  
Routing Information Protocol (RIP)  
Address Resolution Protocol  
(ARP) table  
BOOTP and DHCP clients  
BOOTP relay agent  
Y
Y
Y
Y
Virtual Router Redundancy  
Protocol  
Port Security  
MAC address-based port security  
Y
Y
Y
Y
Y
Y
802.1x port-based network access  
control  
Management Security  
Web server  
Y
Y
Y
Y
Y4  
Y5  
Encryption keys  
Public Key Infrastructure (PKI)  
certificates and Secure Sockets  
Layer (SSL) protocol  
Y
Y
Secure Shell server  
Y
Y
Y
Y
Y
Y
TACACS+ and RADIUS  
authentication  
Management access control list  
Y
Y
Y
1. From the web browser interface you can view the files in the file system of the switch and on a compact flash  
card, but you cannot: copy, rename, or delete them; change directories on a compact flash card; or create a  
new switch configuration file.  
39  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 1: Overview  
2. You cannot upload or download files to a compact flash card with the web browser interface. Also, the inter-  
face does not support switch-to-switch uploads.  
3. You cannot modify the event log full action from the web browser interface.  
4. You can view the encryption keys from the web browser interface, but you cannot create or delete them.  
5. You can view the PKI certificates and the SSL and PKI settings from the web browser interface, but you  
cannot create or delete certificates; create or delete certificate enrollment requests; or change the settings.  
40  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Management Access Methods  
You can access the AT-S63 Management Software on the switch several  
ways:  
ˆ Local session  
ˆ Remote Telnet session  
ˆ Remote Secure Shell (SSH) session  
ˆ Remote web browser (HTTP or HTTPS) session  
ˆ Remote SNMP session  
Local You establish a local management session to the switch by connecting a  
terminal or a PC with a terminal emulator program to the Terminal Port on  
Management  
Sessions  
the front panel using the management cable included with the unit. A local  
management session must be performed at the switch, hence the name  
“local.”  
The switch does not need an Internet Protocol (IP) configuration for local  
management. You can use either the command line interface or the  
menus interface from a local management session. The web browser  
interface is not available from this type of management session.  
Note  
In most cases, the initial management session of a switch must be a  
local management session.  
Remote Telnet The AT-S63 Management Software comes with a Telnet server for remote  
management of the unit from a Telnet client on your network using the  
menus interface or the command line interface.  
Sessions  
Remote Secure Also included in the AT-S63 Management Software is a Secure Shell  
(SSH) server for remote management from a SSH client on your network.  
An SSH management session is similar to a Telnet management session  
except it uses encryption to protect the management sessions from  
Shell (SSH)  
Sessions  
snooping.  
Remote Web The AT-S63 Management Software also comes with a web browser server  
and a web browser interface for remote management using a web browser  
at a workstation on your network. A web browser session can be either  
non-encrypted (HTTP) or encrypted (HTTPS).  
Browser Session  
41  
Download from Www.Somanuals.com. All Manuals Search And Download.  
                 
Chapter 1: Overview  
Remote SNMP You can also remotely configure the switch using a Simple Network  
Management Protocol (SNMP) application, such as AT-View. This  
management method requires an understanding of management  
information base (MIB) objects.  
Management  
The AT-S63 Management Software supports the following MIBs:  
ˆ SNMP MIB-II (RFC 1213)  
ˆ Bridge MIB (RFC 1493)  
ˆ Interface Group MIB (RFC 2863)  
ˆ Ethernet MIB (RFC 1643)  
ˆ Remote Network MIB (RFC 1757)  
ˆ Allied Telesis managed switch MIBs  
The Allied Telesis managed switch MIBs (atistackinfo.mib and  
atistackswitch.mib) are available from the Allied Telesis web site.  
Note  
The switch must have an IP address for remote Telnet, SSH, or  
SNMP management. For background information, refer to “IP  
42  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Manager Access Levels  
The AT-S63 Management Software has two manager access levels of  
manager and operator. The manager access level lets you view and  
configure the operating parameters, while the operator access level only  
lets you only view the parameters settings.  
You log in by entering the appropriate username and password when you  
start a management session. To log in as a manager, type “manager” as  
the login name. The default password is “friend.” The username for  
operator is “operator” and the default password is also “operator.” The  
usernames and passwords are case sensitive.  
There can be only one manager session on a switch at a time. However,  
there can be up to nine simultaneous operator sessions if there is no  
active manager session, or eight operator sessions if there is an active  
manager session.  
43  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 1: Overview  
Installation and Management Configurations  
The AT-9400 Switches can be installed in three configurations.  
Stand-alone All the AT-9400 Switches can be installed and operated as managed or  
unmanaged, stand-alone Gigabit Ethernet switches. Stand-alone switches  
are managed by initiating a local or remote session on the unit.  
Switch  
Enhanced You can simplify the management of the switches in your network by  
connecting them together into an enhanced stack. What this feature does  
Stacking  
is it allows you to quickly and easily transition during a management  
session between the different switches in the network. When you are  
finished managing one switch in an enhanced stack, you can redirect the  
session to another unit without having to end the initial session.  
It is important to note, however, that even through the switches of an  
enhanced stack can be managed from the same management session,  
they operate as independent units, just like stand-alone switches, and are  
configured individually.  
Other highlights to the enhanced stacking feature are:  
ˆ The switches are connected by a common virtual LAN.  
ˆ The devices can be located across a large geographical area.  
ˆ All AT-9400 Switches support this feature.  
For more information, refer to Chapter 2, “Enhanced Stacking” on page  
Stacking Three models in the AT-9400 Basic Layer 3 Series support a third  
installation configuration called stacking. Built with the AT-StackXG  
Stacking Module, a stack merges and synchronizes the network  
operations of two or more AT-9400 Switches to form a single, logical unit  
so that network functions, like the spanning tree protocols, virtual LANs,  
and static port trunks, can span all the Gigabit Ethernet ports of the units in  
the stack.  
There are two principal advantages of a stack over stand-alone switches.  
First, you can configure the switches of a stack simultaneously from the  
same management session, rather than individually from different  
sessions, simplifying management.  
A stack also offers more flexibility in customizing the features of the  
switches for your network. For instance, the ports of a static port trunk on a  
stand-alone switch must be members of the same switch, while the ports  
of a static trunk on a stack can be selected from different switches in the  
same stack.  
44  
Download from Www.Somanuals.com. All Manuals Search And Download.  
         
AT-S63 Management Software Features Guide  
Here are the main points of stacking:  
ˆ The AT-9400 Gigabit Ethernet Switches operate as a single, logical  
unit where functions such as port trunks and port mirrors, can span all  
of the devices in the stack.  
ˆ The switches are managed as a unit.  
ˆ The switches share a common MAC address table.  
ˆ The switches must be installed in the same wiring closet in the same  
equipment rack.  
ˆ The switches are cabled together with the AT-StackXG Stacking  
Module.  
ˆ This stack feature is only supported on the AT-9424Ts, AT-9424Ts/XP,  
and AT-9448Ts/XP Switches.  
For more information on stacking, refer to Chapter 1, Overview in the  
AT-S63 Stack Command Line Interface User’s Guide.  
45  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 1: Overview  
IP Configuration  
Do you intend to remotely manage the switch with a Telnet or Secure  
Shell client, or a web browser? Or, will the management software be  
accessing application servers on your network, like a Simple Network  
Network Time Protocol server for setting its date and time, or a TFTP  
server for uploading or downloading files? If so, then the switch will need  
an IP configuration.  
To assign an IP configuration to the switch, you need to create a routing  
interface. This takes planning because there are number of factors that  
have to be taken into account. For example, you need to know if the switch  
is an AT-9400 Layer 2+ Switch, which supports only one routing interface,  
or an AT-9400 Basic Layer 3 Switch, which supports more than one  
routing interface. If the answer is the latter, you also have to consider  
whether your plans include implementing Internet Protocol version 4  
packet routing on the switch. Furthermore, since routing interfaces are  
assigned to virtual LANs (VLANs), you might need to create one or more  
VLANs on the switch.  
For background information, refer to “Routing Interfaces and Management  
Routing” on page 299. If your plans include implementing IPv4 packet  
routing, you should probably read that entire chapter. For background  
information on VLANs, refer to Chapter 22, “Port-based and Tagged  
46  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Redundant Twisted Pair Ports  
Several AT-9400 Switches have twisted pair ports and GBIC or SFP slots  
that are paired together. The twisted pair ports are identified with the letter  
“R” for “Redundant” as part of their number on the front faceplate of the  
unit. The switch models with paired ports and slots are listed in Table 3.  
Table 3 Twisted Pair Ports Matched with GBIC and  
SFP Slots  
Model  
Ports and Slots  
AT-9424T/GB  
23R with GBIC slot 23  
24R with GBIC slot 24  
AT-9424T/SP  
23R with SFP slot 23  
24R with SFP slot 24  
AT-9424T,  
21R with SFP slot 21  
AT-9424Ts and 22R with SFP slot 22  
AT-9424Ts/XP 23R with SFP slot 23  
24R with SFP slot 24  
AT-9448T/SP  
45R with SFP slot 45  
46R with SFP slot 46  
47R with SFP slot 47  
48R with SFP slot 48  
Follow these guidelines when using these ports and slots:  
ˆ Only one port in a pair — either the twisted pair port or the companion  
GBIC or SFP module — can be active at a time.  
ˆ The twisted pair port is the active port when its GBIC or SFP slot is  
empty, or when a GBIC or SFP module is installed but has not  
established a link to an end node.  
ˆ The twisted pair port automatically changes to the redundant status  
mode when a GBIC or SFP module establishes a link with an end  
node.  
ˆ A twisted pair port automatically transitions back to the active status  
when the link is lost on the GBIC or SFP module.  
ˆ A twisted pair port and a GBIC or SFP module share the same  
configuration settings, including port settings, VLAN assignments,  
access control lists, and spanning tree.  
ˆ The only exception to shared settings is port speed. If you disable  
Auto-Negotiation on a twisted pair port and set the speed and duplex  
mode manually, the speed reverts to Auto-Negotiation when a GBIC or  
SFP module establishes a link with an end node.  
47  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 1: Overview  
Note  
These guidelines do not apply to the SFP slots on the  
AT-9408LC/SP switch and the XFP slots on the AT-9424Ts/XP and  
AT-9448Ts/XP switches.  
48  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
History of New Features  
The following sections contain the history of new features in the AT-S63  
Management Software.  
Version 3.0.0 Table 4 lists the new features in version 3.0.0 of the AT-S63 Management  
Software.  
Table 4. New Features in AT-S63 Version 3.0.0  
Feature  
Change  
Stacking with the AT-StackXG  
Stacking Module  
New feature. For information, refer  
to Chapter 1, Overview in the  
AT-S63 Stack Command Line  
Interface User’s Guide.  
Virtual Router Redundancy  
Protocol (VRRP)  
New feature. For information, refer  
337.  
Ethernet Protection Switching  
Ring (EPSR) snooping  
New feature. For information, refer  
Internet Protocol version 4 packet  
routing  
Added the following new features:  
ˆ Split horizon with poison  
reverse  
ˆ Auto-summarization of routes  
ˆ DHCP/BOOTP relay  
802.1x port-based network access Added the following authentication  
control  
methods:  
ˆ EAP-TLS (Extensible  
Authentication Protocol -  
Transport Layer Security)  
ˆ EAP-TTLS (Extensible  
Authentication Protocol -  
Tunneled Transport Layer  
Security)  
ˆ PEAP (Protected Extensible  
Authentication Protocol)  
49  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 1: Overview  
Version 2.1.0 Table 5 lists the new features in version 2.1.0.  
Table 5. New Features in AT-S63 Version 2.1.0  
Feature  
Change  
Internet Protocol version 4 packet  
routing  
Added the following new features:  
ˆ Equal Cost Multi-path (ECMP)  
for supporting multiple routes  
in the routing table to the same  
remote destination.  
ˆ Variable length subnet masks  
for the IP addresses of routing  
interfaces and static and  
dynamic routes.  
Version 2.0.0 Table 6 lists the new feature in version 2.0.0 of the AT-S63 Management  
Software.  
Table 6. New Features in AT-S63 Version 2.0.0  
Feature  
Change  
Internet Protocol version 4 packet  
routing with:  
New feature.  
ˆ Routing interfaces  
ˆ Static routes  
ˆ Router Information Protocol  
(RIP) versions 1 and 2  
50  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
AT-S63 Management Software Features Guide  
Version 1.3.0 Table 7 lists the new features in version 1.3.0 of the AT-S63 Management  
Software.  
Table 7. New Features in AT-S63 Version 1.3.0  
Feature  
Change  
802.1x Port-based Network  
Access Control  
Added the following new features:  
ˆ Guest VLAN. For background  
ˆ VLAN Assignment and Secure  
VLAN for supporting dynamic  
VLAN assignments from a  
RADIUS authentication server  
for supplicant accounts. For  
background information, see  
ˆ MAC address-based  
authentication as an  
alternative to 802.1x username  
and password authentication.  
For background information,  
Management Access Control List  
Simplified the menu interface for  
managing the access control  
entries in the Management ACL.  
51  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 1: Overview  
Version 1.2.0 Table 8 lists the new features in version 1.2.0.  
Table 8. New Features in AT-S63 Version 1.2.0  
Feature  
Change  
MAC Address Table  
Added the following new  
parameters to the CLI commands  
for displaying and deleting specific  
types of MAC addresses in the  
MAC address table:  
ˆ STATIC, STATICUNICAST,  
and, STATICMULTICAST for  
displaying and deleting static  
unicast and multicast MAC  
addresses.  
ˆ DYNAMIC,  
DYNAMICUNICAST, and,  
DYNAMICMULTICAST for  
displaying and deleting  
dynamic unicast and multicast  
MAC addresses.  
Quality of Service  
Added the following new  
parameters to QoS flow groups,  
traffic classes, and policies:  
ˆ ToS parameter for replacing  
the Type of Service field of  
IPv4 packets.  
ˆ Move ToS to Priority  
parameter for replacing the  
value in the 802.1p priority  
field with the value in the ToS  
priority field in IPv4 packets.  
ˆ Move Priority to ToS  
parameter for replacing the  
value in the ToS priority field  
with the 802.1p priority field in  
IPv4 packets.  
ˆ Send to Mirror Port parameter  
for copying traffic to a  
destination mirror port (policies  
only)  
MLD Snooping  
New feature.  
New feature.  
MAC Address-based VLANs  
52  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Table 8. New Features in AT-S63 Version 1.2.0 (Continued)  
Feature  
Change  
802.1x Port-based Network  
Access Control  
Added a new parameter to  
authenticator ports:  
ˆ Supplicant Mode for  
supporting multiple supplicant  
accounts on an authenticator  
port. For background  
53  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 1: Overview  
54  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 2: Enhanced Stacking  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three management interfaces in the  
AT-S63 Management Software:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
56  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
Having to manage a large number of network devices typically involves  
starting a separate management session on each device. This usually  
means having to end one management session in order to start a new  
session on another unit.  
The enhanced stacking feature can simplify this task because it allows you  
to easily transition among the different AT-9400 Switches in your network  
from just one management session. This reduces the need of having to  
end a management session when you need to manage another switch.  
It should be noted that the individual switches of an enhanced stack  
function autonomously. They do not form what is commonly referred to as  
a “virtual stack,” where the switches act as a logical unit. Rather, each  
switch in an enhanced stack functions independently of the others.  
Note  
Starting with version 2.0.0 of the AT-S63 Management Software,  
several significant changes have been made to the implementation  
of the enhanced stacking feature. Allied Telesis recommends  
reviewing the information in this section before using this feature,  
even if you are familiar with it from earlier versions of the AT-S63  
Management Software or from other Allied Telesis Ethernet  
switches that support this feature.  
Section I: Basic Operations  
57  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 2: Enhanced Stacking  
Master and Slave Switches  
An enhanced stack must have at least one master switch. This switch is  
your management access point to the switches of a stack. After you have  
started a local or remote management session on a master switch, you  
can redirect the session to any of the other switches.  
The other switches in the stack are known as slave switches. They can be  
managed through the master switch or directly, such as from a local  
management session.  
An enhanced stack can have more than one master switch. Multiple  
master switches can lessen the impact on your network management  
should you need to remove a master switch from the network, such as for  
maintenance purposes.  
58  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Common VLAN  
A master switch searches for the other switches in an enhanced stack by  
sending out a broadcast packet out a local subnet. (The designation of this  
Since a broadcast packet cannot cross a router or a VLAN boundary, you  
must connect the switches of an enhanced stack with a common VLAN.  
The VLAN acts as the transfer path for the broadcast packets from the  
master switch to the slave switches and also serves as the path for other  
management packets.  
Here are several things to keep in mind as you plan the common VLAN of  
your enhanced stack:  
ˆ Any valid VLAN name and VLAN identifier (VID) can be used for the  
common VLAN, but it should be the same on all the switches in the  
stack.  
ˆ A slave switch of an enhanced stack can be indirectly connected to the  
master switch through other switches, provided there is an  
uninterrupted path of the common VLAN from the slave switch to the  
master switch.  
ˆ The Default_VLAN can be used as the common VLAN.  
ˆ The common VLAN does not have to be dedicated solely to the  
enhanced stacking feature.  
For background information on port-based and tagged virtual LANs, refer  
Section I: Basic Operations  
59  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 2: Enhanced Stacking  
Master Switch and the Local Interface  
Before a switch can function as the master switch of an enhanced stack, it  
needs to know which subnet is acting as the common subnet among the  
switches in the stack. It uses that information to know which subnet to  
send out its broadcast packets and to monitor for the management  
packets from the other switches and from remote management  
workstations.  
Designating the common VLAN and subnet involves creating a routing  
interface on the master switch on the common subnet and designating it  
as the local interface. The concept of routing interfaces first appeared in  
the AT-9400 Switch with Layer 3 routing and the implementation of static  
routing and the Routing Information Protocol (RIP) version 1 and 2.  
An interface represents a logical connection to a network or subnet local to  
the switch for purposes of routing packets. To configure an interface, you  
assign it an IP address and subnet mask appropriate to the subnet where  
it will route packets, and add it to the VLAN that contains the subnet.  
For the most part, routing interfaces are limited to the IPv4 packet routing  
feature and are unnecessary beyond that feature. There are, however, a  
few exceptions. One is the enhanced stacking feature. The rule is that the  
master switch of an enhanced stack must have at least one interface and  
the interface must be assigned to the common subnet that interconnects  
the switches of the stack. Furthermore, the interface must be designated  
as the switch’s local interface. The act of designating an interface as the  
local interface tells the switch which interface and which subnet it should  
use for the enhanced stacking feature.  
For background information on the IPv4 routing feature, refer Chapter 27,  
60  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Slave Switches  
The slave switches of an enhanced stack must be connected to the master  
switch through a common VLAN. A slave switch can be connected  
indirectly to the master switch so long as there is an uninterrupted path of  
the common VLAN from the slave switch to the master switch.  
A slave switch does not need a routing interface on the common VLAN if  
you use the Default_VLAN (VID 1) as the common VLAN. A routing  
interface in the common VLAN is required if you use any other VLAN other  
than the Default_VLAN as the common VLAN of the switches in the stack.  
The routing interface in the common VLAN on a slave switch does not  
have to be designated as the local interface. The only circumstance in  
which you might want to designate a local interface on a slave switch is if  
you want to be able to remotely manage the device independently of the  
enhanced stack. However, for the switch to remain part of an enhanced  
stack, the interface designated as the local interface must be in the  
common VLAN.  
Section I: Basic Operations  
61  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 2: Enhanced Stacking  
Enhanced Stacking Compatibility  
This version of enhanced stacking is compatible with earlier AT-S63  
versions and the enhanced stacking feature in the AT-8000 Series,  
AT-8400 Series, and AT-8500 Series Switches. As such, an enhanced  
stack can consist of various switch models, though the following issues  
need to be considered when building this type of enhanced stack:  
ˆ The management VLAN of an AT-8000 Series, AT-8400 Series, or  
AT-8500 Series Switch must be assigned to the common VLAN that  
interconnects the switches of the stack. For instructions on how to  
select the management VLAN on an AT-8000 Series, AT-8400 Series,  
or AT-8500 Series switch, refer to the appropriate user’s guide.  
ˆ Though the master switch of an enhanced stack can be any switch  
that supports this feature, Allied Telesis recommends choosing the  
AT-9400 Switch to perform that role. To use an AT-8000 Series,  
AT-8400 Series, or AT-8500 Series switch as the master switch, you  
must assign it an IP address that is part of the same common subnet  
that interconnects the switches of the stack. For instructions on how to  
assign an IP address to an AT-8000 Series, AT-8400 Series, or  
AT-8500 Series switch, refer to the appropriate user’s guide.  
62  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Enhanced Stacking Guidelines  
Here are the guidelines to using the enhanced stacking feature:  
ˆ There can be up to 24 switches in an enhanced stack.  
ˆ The switches in an enhanced stack must be connected with a common  
port-based or tagged VLAN. The VLAN must have the same name and  
VLAN identifier (VID) on each switch, and the switches must be  
connected using tagged or untagged ports of the VLAN.  
ˆ A slave switch can be connected indirectly to the master switch  
through other switches so long as there is an uninterrupted path of the  
common VLAN from the master switch to the slave switch.  
ˆ You must add a routing interface to the common VLAN on the master  
switch and designate it as the master switch’s local interface.  
ˆ You do not need to create a routing interface in the common VLAN on  
the slave switches if you use the Default_VLAN (VID 1) as the  
common VLAN of the switches of a stack. However, a routing interface  
is required if you use any other VLAN as the common VLAN. However,  
you do not have to designate it as the local interface.  
ˆ You can create different stacks by connecting different groups of  
switches with different common VLANs and subnets.  
ˆ An enhanced stack must have at least one master switch. You  
designate the master by changing its stacking status to Master.  
ˆ An enhanced stack can consist of other Allied Telesis switches that  
support this feature, including the AT-8000, AT-8400, AT-8500, and  
AT-9400 Switches. For more information, refer to “Enhanced Stacking  
ˆ In order to manage the stack remotely using a Telnet or SSH client or a  
web browser, the remote management workstation must reach the  
master switch through the subnet of the switch’s local interface.  
ˆ The IP address 172.16.16.16 is reserved for the enhanced stacking  
feature and must not be assigned to any device on your network.  
Section I: Basic Operations  
63  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 2: Enhanced Stacking  
General Steps  
Here are the basic steps to implementing the enhanced stacking feature  
on the AT-9400 Switches in your network:  
1. Select a switch to act as the master switch of the enhanced stack. This  
can be any Allied Telesis switch that supports this feature. In a stack  
with different switch models, Allied Telesis recommends using an  
AT-9400 Switch as the master switch. For further information, refer to  
2. On the switch chosen to be the master switch, change its stacking  
status to Master.  
3. Create a common port-based or tagged VLAN on each switch and  
connect the devices using twisted pair or fiber optic ports of the VLAN.  
As mentioned earlier, the slaves switches can be connected indirectly  
through other switches to the master switch, so long as there is an  
uninterrupted path of the common VLAN to the master switch. This  
step is not necessary if you use the Default_VLAN (VID 1) as the  
common VLAN.  
4. On the master switch, assign a routing interface to the common VLAN.  
5. On the master switch designate the interface assigned to the common  
VLAN as the local interface.  
6. On the slave switches, add a routing interface to the common VLAN.  
You do not need to designate it as the local interface. This step is not  
necessary if you use the Default_VLAN (VID 1) as the common VLAN.  
Note  
The initial configuration of the enhanced stacking feature on a  
master switch must be performed through a local management  
session.  
64  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 3  
SNMPv1 and SNMPv2c  
This chapter describes SNMPv1 and SNMPv2c community strings for  
SNMP management of the switch. Sections in the chapter include:  
Section I: Basic Operations  
65  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 3: SNMPv1 and SNMPv2c  
Supported Platforms  
This feature is supported on all AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three management interfaces in the  
AT-S63 Management Software:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
66  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
You can manage a switch by viewing and changing the management  
information base (MIB) objects on the device with the Simple Network  
Management Program (SNMP). The AT-S63 Management Software  
supports SNMPv1, SNMPv2c, and SNMPv3. This chapter explains  
SNMPv1 and SNMPv2c. For information on SNMPv3, refer to Chapter 19,  
To manage a switch using an SNMP application program, you must do the  
following:  
ˆ Activate SNMP management on the switch. The default setting for  
SNMP management is disabled.  
ˆ Load the Allied Telesis MIBs for the switch onto your management  
workstation containing the SNMP application program. The MIBs are  
available from the Allied Telesis web site at www.alliedtelesis.com.  
To manage a switch using SNMP, you need to know the IP address of the  
switch or of the master switch of an enhanced stack and at least one of the  
switch’s community strings.  
You can configure SNMPv1 and SNMPv2c with the SNMPv3 Table menus  
Table menus require a much more extensive configuration.  
Section I: Basic Operations  
67  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 3: SNMPv1 and SNMPv2c  
Community String Attributes  
A community string has attributes for controlling who can use the string  
and what the string will allow a network management to do on the switch.  
The community string attributes are defined below:  
Community A community string must have a name of one to eight alphanumeric  
characters. Spaces are allowed.  
String Name  
Access Mode This attribute defines the permissions of a community string. There are  
two access modes: Read and Read/Write. A community string with an  
access mode of Read can only be used to view but not change the MIB  
objects on a switch. A community string with a Read/Write access can be  
used to both view the MIB objects and change them.  
Operating Status A community string can be enabled or disabled. When disabled, no one  
can use it to access the switch. You might disable a community string if  
you suspect someone is using it for unauthorized access to the device.  
When a community string is enabled, then it is available for use.  
Open or Closed This feature controls which management stations on your network can use  
a community string. An open access status permits any network manager  
Access Status  
who knows the community string to use it. A closed access status restricts  
the string to those network managers who work at particular workstations,  
identified by their IP addresses. You specify the workstations by assigning  
the IP addresses of the workstations to the community string. A closed  
community string can have up to eight IP addresses of management  
workstations.  
If you decide to activate SNMP management on the switch, it is a good  
idea to assign a closed status to all community strings that have a Read/  
Write access mode and then assign the IP addresses of your  
management workstations to those strings. This helps reduce the chance  
of someone gaining management access to a switch through a community  
string and making unauthorized configuration changes.  
Trap Receivers A trap is a signal sent to one or more management workstations by the  
switch to indicate the occurrence of a particular operating event on the  
device. There are numerous operating events that can trigger a trap. For  
instance, resetting the switch or the failure of a cooling fan are two  
examples of occurrences that cause a switch to send a trap to the  
management workstations. You can use traps to monitor activities on the  
switch.  
Trap receivers are the devices, typically management workstations or  
servers, that you want to receive the traps sent by the switch. You specify  
the trap receivers by their IP addresses. You assign the IP addresses to  
68  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
                     
AT-S63 Management Software Features Guide  
the community strings.  
Each community string can have up to eight trap IP addresses.  
It does not matter which community strings you assign your trap receivers.  
When the switch sends a trap, it looks at all the community strings and  
sends the trap to all trap receivers on all community strings. This is true  
even for community strings that have a access mode of only Read.  
If you are not interested in receiving traps, then you do not need to enter  
any IP addresses of trap receivers.  
Section I: Basic Operations  
69  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 3: SNMPv1 and SNMPv2c  
Default SNMP Community Strings  
The AT-S63 Management Software provides two default community  
strings: public and private. The public string has an access mode of just  
Read and the private string has an access mode of Read/Write. If you  
activate SNMP management on the switch, you should delete or disable  
the private community string, which is a standard community string in the  
industry, or change its status from open to closed to prevent unauthorized  
changes to the switch.  
70  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 4  
MAC Address Table  
This chapter contains background information about the MAC address  
table.This chapter contains the following section:  
Section I: Basic Operations  
71  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 4: MAC Address Table  
Overview  
The AT-9400 Switch has a MAC address table with a storage capacity of  
16,000 entries. The table stores the MAC addresses of the network nodes  
connected to its ports and the port number where each address was  
learned.  
The switch learns the MAC addresses of the end nodes by examining the  
source address of each packet received on a port. It adds the address and  
port on which the packet was received to the MAC table if the address has  
not already been entered in the table. The result is a table that contains all  
the MAC addresses of the devices that are connected to the switch’s  
ports.  
When the switch receives a packet, it also examines the destination  
address and, by referring to its MAC address table, determines the port  
where the destination node is connected. It then forwards the packet to the  
appropriate port and on to the end node. This increases network  
bandwidth by limiting each frame to the appropriate port when the  
intended end node is located, freeing the other switch ports for receiving  
and transmitting other packets.  
If the switch receives a packet with a destination address that is not in the  
MAC address table, it floods the packet to all the ports on the switch,  
excluding the port where the packet was received. If the ports have been  
grouped into virtual LANs, the switch floods the packet only to those ports  
that belong to the same VLAN from where the packet originated. This  
prevents packets from being forwarded onto inappropriate LAN segments  
and increases network security. When the destination node responds, the  
switch adds its MAC address and port number to the table.  
If the switch receives a packet with a destination address that is on the  
same port on which the packet was received, it discards the packet  
without forwarding it on to any port. Because both the source node and the  
destination node for the packet are located on the same port on the  
switch, there is no reason for the switch to forward the packet. This too  
increases network performance by preventing frames from being  
forwarded unnecessarily to other network devices.  
The type of MAC address described above is referred to as a dynamic  
MAC address. Dynamic MAC addresses are addresses that the switch  
learns by examining the source MAC addresses of the frames received on  
the ports.  
Dynamic MAC addresses are not stored indefinitely in the MAC address  
table. The switch deletes a dynamic MAC address from the table if it does  
not receive any frames from the node after a specified period of time. The  
switch assumes that the node with that MAC address is no longer active  
and that its MAC address can be purged from the table. This prevents the  
72  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
AT-S63 Management Software Features Guide  
MAC address table from becoming filled with addresses of nodes that are  
no longer active.  
The period of time that the switch waits before purging an inactive dynamic  
MAC address is called the aging time. This value is adjustable on the  
AT-9400 Switch. The default value is 300 seconds (5 minutes).  
The MAC address table can also store static MAC addresses. A static  
MAC address is a MAC address of an end node that you assign to a switch  
port manually. A static MAC address remains in the table indefinitely and  
is never deleted, even when the end node is inactive.  
You might need to enter static MAC addresses of end nodes the switch  
does not learn in its normal dynamic learning process, or if you want a  
MAC address to remain permanently in the table, even when the end node  
is inactive.  
Section I: Basic Operations  
73  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 4: MAC Address Table  
74  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 5  
Static Port Trunks  
This chapter describes static port trunks. Sections in the chapter include:  
Section I: Basic Operations  
75  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 5: Static Port Trunks  
Supported Platforms  
This feature is supported on all AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
76  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
A static port trunk is a group of two to eight ports that function as a single  
virtual link between the switch and another device. Traffic is distributed  
across the ports to improve performance and enhance reliability by  
reducing the reliance on a single physical link.  
A static port trunk is easy to configure. You simply designate the ports of  
the trunk and the management software automatically groups them  
together. You can also control how traffic is distributed over the trunk  
example in Figure 1 illustrates a static port trunk of four links between two  
AT-9400 Switches.  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
8
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
23  
24  
POWER  
2
4
10 12 14 16 18 20 22 24R  
2
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
D/C  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
23  
24  
POWER  
2
4
6
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Figure 1. Static Port Trunk Example  
Redundancy and link backup are not supported in a static trunk. If a link is  
lost on a port, the trunk’s total bandwidth is reduced. Although the traffic  
carried by the lost link is shifted to one of the remaining ports in the trunk,  
the bandwidth remains reduced until the lost link is reestablished or  
another port is added to the trunk.  
Network equipment vendors tend to employ different techniques for static  
trunks on their products. Consequently, a static trunk on one device might  
not be compatible with the same feature on a device from a different  
manufacturer. For this reason, static trunks are typically employed only  
between devices from the same vendor.  
Section I: Basic Operations  
77  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 5: Static Port Trunks  
Load Distribution Methods  
This section discusses load distribution methods and applies to both static  
and LACP port trunks.  
One of the steps to creating a static or LACP port trunk is selecting a load  
distribution method, which determines how the switch distributes the traffic  
load across the ports in the trunk. The AT-S63 Management Software  
offers the following load distribution methods:  
ˆ Source MAC Address (Layer 2)  
ˆ Destination MAC Address (Layer 2)  
ˆ Source MAC Address / Destination MAC Address (Layer 2)  
ˆ Source IP Address (Layer 3)  
ˆ Destination IP Address (Layer 3)  
ˆ Source IP Address / Destination IP Address (Layer 3)  
The load distribution methods examine the last three bits of a packet’s  
MAC or IP address and compare the bits against mappings assigned to  
the ports in the trunk. The port mapped to the matching bits is selected as  
the transmission port for the packet.  
In cases where you select a load distribution that employs either a source  
or destination address but not both, only the last three bits of the  
designated address are used in selecting a transmission port in a trunk. If  
you select one of the two load distribution methods that employs both  
source and destination addresses, port selection is achieved through an  
XOR operation of the last three bits of both addresses.  
As an example, assume you created a static or LACP aggregate trunk of  
Ports 7 to 14 on a switch. The table below shows the mappings of the  
switch ports to the possible values of the last three bits of a MAC or IP  
address.  
Last 3 Bits  
000  
(0)  
001  
(1)  
010  
(2)  
011  
(3)  
100  
(4)  
101  
(5)  
110  
(6)  
111  
(7)  
Trunk Ports  
7
8
9
10  
11  
12  
13  
14  
Assume you selected source MAC address as the load distribution  
method and that the switch needed to transmit over the trunk a packet with  
a source MAC address that ended in 9. The binary equivalent of 9 is 1001,  
making the last three bits of the address 001. An examination of the table  
above indicates that the switch would use Port 8 to transmit the frame  
because that port is mapped to the matching bits.  
78  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
AT-S63 Management Software Features Guide  
A similar method is used for the two load distribution methods that employ  
both the source and destination addresses. Only here the last three bits of  
both addresses are combined by an XOR process to derive a single value  
which is then compared against the mappings of the bits to ports. The  
XOR rules are as follows:  
0 XOR 0 = 0  
0 XOR 1 = 1  
1 XOR 0 = 1  
1 XOR 1 = 0  
As an example, assume you had selected source and destination MAC  
addresses for the load distribution method in our previous example, and  
that a packet for transmission over the trunk had a source MAC address  
that ended in 9 and a destination address that ended in 3. The binary  
values would be:  
9 = 1001  
3 = 0011  
Applying the XOR rules above on the last three bits would result in 010, or  
2. A examination of the table above shows that the packet would be  
transmitted from port 9.  
Port trunk mappings on the AT-9400 Switch can consist of up to eight  
ports. This corresponds to the maximum number of ports allowed in a  
static trunk and the maximum number of active ports in an LACP trunk.  
Inactive ports in an LACP trunk are not applied to the mappings until they  
transition to the active status.  
You can assign different load distribution methods to different static trunks  
on the same switch. The same is true for LACP aggregators. However, it  
should be noted that all aggregate trunks within an LACP aggregator must  
use the same load distribution method.  
The load distribution methods assume that the final three bits of the source  
and/or destination addresses of the packets from the network nodes are  
varied enough to support efficient distribution of the packets over the trunk  
ports. A lack of variation can result in one or more ports in a trunk being  
used more than others, with the potential loss of a trunk’s efficiency and  
performance.  
Section I: Basic Operations  
79  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 5: Static Port Trunks  
Guidelines  
The following guidelines apply to static trunks:  
ˆ Allied Telesis recommends limiting static port trunks to Allied Telesis  
network devices to ensure compatibility.  
ˆ A static trunk can have up to eight ports.  
ˆ Stand-alone switches can support up to six static and LACP trunks at a  
time (for example, four static trunks and two LACP trunks). An LACP  
trunk is countered against the maximum number of trunks only when it  
is active.  
ˆ Stacks of AT-9400 Basic Layer 3 Switches and the AT-StackXG  
Stacking Module can support up to six static port trunks.  
ˆ The ports of a static trunk must be of the same type of either twisted  
pair or fiber optic ports.  
ˆ The ports of a trunk can be either consecutive (for example Ports 5-9)  
or nonconsecutive (for example, ports 4, 8, 11, 20).  
ˆ The ports of static port trunks on stand-alone switches or switches in  
an enhanced stack must be from the same switch.  
ˆ The ports of a static port trunk in a stack of AT-9400 Basic Layer 3  
Switches and the AT-StackXG Stacking Module can be from different  
switches in the same stack.  
ˆ Before creating a port trunk, examine the speed, duplex mode, flow  
control, and back pressure settings of the lowest number port to be in  
the trunk. Verify that its settings are correct for the device to which the  
trunk will be connected. When you create a static port trunk, the  
management software copies the current settings of the lowest  
numbered port in the trunk to the other ports, because all ports in a  
static trunk must have the same settings. For example, if you create a  
port trunk consisting of ports 5 to 8, the parameter settings for port 5  
are copied to ports 6, 7, and 8 so that all the ports of the trunk have the  
same settings.  
ˆ After creating a port trunk, do not change the speed, duplex mode,  
flow control, or back pressure of any port in the trunk without also  
changing the other ports.  
ˆ A port can belong to only one static trunk at a time.  
ˆ A port cannot be a member of a static trunk and an LACP trunk at the  
same time.  
ˆ The ports of a static trunk must be untagged members of the same  
VLAN. A trunk cannot consist of untagged ports from different VLANs.  
ˆ The switch selects the lowest numbered port in the trunk to handle  
broadcast packets and packets of unknown destination. For example,  
a trunk of ports 11 to 15 would use port 11 for broadcast packets.  
80  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 6  
LACP Port Trunks  
This chapter explains Link Aggregation Control Protocol (LACP) port  
trunks. Sections in the chapter include:  
Section I: Basic Operations  
81  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 6: LACP Port Trunks  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from two of the management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
82  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Overview  
LACP (Link Aggregation Control Protocol) port trunks perform the same  
function as static trunks. They increase the bandwidth between network  
devices by distributing the traffic load over multiple physical links. The  
advantage of an LACP trunk over a static port trunk is its flexibility. While  
implementations of static trunking tend to be vendor specific, the  
implementation of LACP in the AT-S63 Management Software is  
compliant with the IEEE 802.3ad standard, making it interoperable with  
equipment from other vendors that also comply with the standard.  
Therefore, you can create an LACP trunk between an Allied Telesis device  
and network devices from other manufacturers.  
Another advantage is that ports in an LACP trunk can function in a standby  
mode. This adds redundancy and resiliency to the trunk. If a link in a static  
trunk goes down, the overall bandwidth of the trunk is reduced until the link  
is reestablished or another port is added to the trunk. In contrast, an LACP  
trunk can automatically activate ports in a standby mode when an active  
link fails so that the maximum possible bandwidth of the trunk is  
maintained.  
For example, assume you create an LACP trunk of ports 11 to 20 on a  
switch and the switch is using ports 11 to 18 as the active ports and ports  
19 and 20 as reserve. If an active port loses its link, the switch  
automatically activates one of the reserve ports to maintain maximum  
bandwidth of the trunk.  
The main component of an LACP trunk is an aggregator. An aggregator is  
a group of ports on the switch. The ports in an aggregator are further  
grouped into one or more trunks, referred to as aggregate trunks.  
An aggregate trunk can consist of any number of ports on a switch, but  
only a maximum of eight ports can be active at a time. If an aggregate  
trunk contains more ports than can be active at one time, the extra ports  
are placed in a standby mode. Ports in the standby mode do not pass  
network traffic, but they do transmit and accept LACP data unit (LACPDU)  
packets, which the switch uses to search for LACP-compliant devices.  
Only ports on a switch that are part of an aggregator transmit LACPDU  
packets. If a switch port that is part of an aggregator does not receive  
LACPDU packets from its corresponding port on the other device, it  
assumes that the other port is not part of an LACP trunk. Instead, it  
functions as a normal Ethernet port by forwarding network traffic.  
However, it does continue to send LACPDU packets. If it begins to receive  
LACPDU packets, it automatically transitions to an active or standby mode  
as part of an aggregate trunk.  
Section I: Basic Operations  
83  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 6: LACP Port Trunks  
If there will be more than one aggregate trunk on a switch, each trunk  
might require a separate aggregator or it might be possible to combine  
them into a common aggregator. The determining factor will be whether  
the trunks are going to the same device or different devices. If the trunks  
are going to the same device, each must have its own aggregator. If they  
are going to different devices, the trunks can be members of a common  
aggregator. In the latter situation, the switch will differentiate the individual  
aggregate trunks.  
Here are two examples. Figure 2 illustrates the AT-9400 Switch with two  
LACP trunks, each containing three links. Because both aggregate trunks  
go to the same 802.3ad-compliant device, in this case another Gigabit  
Ethernet switch, each trunk requires a separate aggregator.  
Ports 12 -14  
in Aggregator 2  
Ports 1 - 3  
in Aggregator 1  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
AT-9400 Switch  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Aggregate Trunks  
in Different Aggregators  
AT-8524M Fast Ethernet Switch  
STATUS  
802.3ad-compliant  
Ethernet Switch  
MODE  
LINK  
FAULT  
MASTER  
RPS  
MODE  
LINK  
MODE  
PWR  
Figure 2. Example of Multiple Aggregators for Multiple Aggregate Trunks  
84  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Here is how the example looks in a table format.  
Aggregator  
Description  
Aggregator  
Ports  
Aggregate  
Trunk Ports  
Aggregator 1  
Aggregator 2  
1-3  
1-3  
12-14  
12-14  
Caution  
The example cited here illustrates a loop in a network. Avoid  
network loops to prevent broadcast storms.  
If the aggregate trunks go to different devices, you can create one  
aggregator and the AT-9400 Switch will form the trunks for you  
automatically. This is illustrated in Figure 3 where the ports of two  
aggregate trunks on the AT-9400 Switch are members of the same  
aggregator. It is the switch that determines that there are actually two  
separate aggregate trunks.  
Ports 1 - 3 and 12-14  
in Aggregator 1  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
AT-9400 Switch  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
4
6
8
10  
12  
14  
16  
1
20  
22  
24R  
Aggregate Trunks  
in Common Aggregator  
AT-8524M Fast Ethernet Switch  
STATUS  
802.3ad-compliant  
Ethernet Switch  
MODE  
802.3ad-compliant  
Server  
LINK  
FAULT  
MASTER  
RPS  
MODE  
LINK  
MODE  
PWR  
Figure 3. Example of an Aggregator with Multiple Trunks  
Section I: Basic Operations  
85  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 6: LACP Port Trunks  
Here is how this example looks in table format.  
Aggregator  
Description  
Aggregator  
Ports  
Aggregate  
Trunk Ports  
Aggregator 1  
1-3, 12-14  
1-3  
12-14  
You could, if you wanted, create separate aggregators for the different  
aggregate trunks in the example above. But letting the switch make the  
determination for you whenever possible saves time later if you physically  
reassign ports to a different trunk connected to another device.  
86  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
LACP System Priority  
It is possible for two devices interconnected by an aggregate trunk to  
encounter a conflict when they form the trunk. For example, the two  
devices might not support the same number of active ports in an  
aggregate trunk or might not agree on which ports are to be active and  
which are to be in standby.  
If a conflict does occur, the two devices need a mechanism for resolving  
the problem and deciding whose LACP settings are to take precedence.  
This is the function of the system LACP priority value. A hexadecimal  
value of from 1 to FFFF, this parameter is used whenever the devices  
encounter a conflict creating a trunk. The lower the number, the higher the  
priority. The settings on the device with the higher priority take precedence  
over the settings on the other device. If both devices have the same  
system LACP priority value, the settings on the switch with the lowest  
MAC address take precedence.  
This parameter can prove useful when connecting an aggregate trunk  
between the AT-9400 Switch and another 802.3ad-compliant device that  
does not have the same LACP trunking capabilities. If the other device’s  
capability is less than that of the AT-9400 Switch, you should give that  
device the higher priority so its settings are used by both devices when  
forming the trunk.  
For example, an aggregate trunk of six links between an AT-9400 Switch  
and an 802.3ad-compliant device that supported up to four active links at  
one time could possibly result in a conflict. The AT-9400 Switch would try  
to use all six links as active, because it can handle up to eight active links  
in a trunk at one time, while the other device would want to use only four  
ports as active. By giving the other 802.3ad device the higher priority, the  
conflict is avoided because the AT-9400 Switch would use only four active  
links, as directed by the other 802.3ad-compliant device. The other ports  
would remain in the standby mode.  
Section I: Basic Operations  
87  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 6: LACP Port Trunks  
Adminkey Parameter  
The adminkey is a hexadecimal value from 1 to FFFF that identifies an  
aggregator. Each aggregator on a switch must have a unique adminkey.  
The adminkey is restricted to a switch. Two aggregators on different  
switches can have the same adminkey without generating a conflict.  
LACP Port Priority Value  
The switch uses a port’s LACP priority to determine which ports are to be  
active and which in the standby mode in situations where the number of  
ports in the aggregate trunk exceeds the highest allowed number of active  
ports. This parameter is a hexadecimal value in a range of 1 to FFFF,  
based on the port number. For instance, the priority values for ports 2 and  
11 are 0002 and 000B, respectively. The lower the number, the higher the  
priority. Ports with the highest priorities are designated as the active ports  
in an aggregate trunk.  
For example, if both 802.3ad-compliant devices support up to eight active  
ports and there are a total of ten ports in the trunk, the eight ports with the  
highest priorities (lowest priority values) are designated as the active  
ports, and the others are placed in the standby mode. If an active link goes  
down on a active port, the standby port with the next highest priority is  
automatically activated to take its place.  
The selection of the active links in an aggregate trunk is dynamic and will  
change as links are added, removed, lost or reestablished. For example, if  
an active port loses its link and is replaced by another port in the standby  
mode, the reestablishment of the link on the originally active port causes  
the port to return to the active state by virtue of having a higher priority  
value than the replacement port, which returns to the standby mode.  
A port’s priority value is not adjustable.  
Two conditions must be met in order for a port in an aggregate trunk to  
function in the standby mode. First, the number of ports in the trunk must  
exceed the highest allowed number of active ports and, second, the port  
must be receiving LACPDU packets from the other device. A port  
functioning in the standby mode does not forward network traffic, but does  
continue to send LACPDU packets. If a port that is part of an aggregator  
does not receive LACPDU packets, it functions as a normal Ethernet port  
and forwards network packets along with LACPDU packets.  
88  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
AT-S63 Management Software Features Guide  
Load Distribution Methods  
The load distribution method determines the manner in which the switch  
distributes the traffic across the active ports of an aggregate trunk. The  
method is assigned to an aggregator and applies to all aggregate trunks  
within it. If you want to assign different load distribution methods to  
different aggregate trunks, you must create a separate aggregator for  
each trunk. For further information, refer to “Load Distribution Methods” on  
Section I: Basic Operations  
89  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 6: LACP Port Trunks  
Guidelines  
The following guidelines apply to creating aggregators:  
ˆ LACP must be activated on both the switch and the other device.  
ˆ The other device must be 802.3ad-compliant.  
ˆ An aggregator can consist of any number of ports.  
ˆ The AT-S63 Management Software supports up to eight active ports in  
an aggregate trunk at a time.  
ˆ The AT-9400 Switch can support up to six static and LACP aggregate  
trunks at a time (for example, four static trunks and two LACP trunks).  
An LACP trunk is countered against the maximum number of trunks  
only when it is active.  
ˆ The ports of an aggregate trunk must be the same medium type: all  
twisted pair ports or all fiber optic ports.  
ˆ The ports of a trunk can be consecutive (for example ports 5-9) or  
nonconsecutive (for example, ports 4, 8, 11, 20).  
ˆ A port can belong to only one aggregator at a time.  
ˆ A port cannot be a member of an aggregator and a static trunk at the  
same time.  
ˆ The ports of an aggregate trunk must be untagged members of the  
same VLAN.  
ˆ 10/100/1000Base-TX twisted pair ports must be set to Auto-  
Negotiation or 100 Mbps, full-duplex mode. LACP trunking is not  
supported in half-duplex mode.  
ˆ 100Base-FX fiber optic ports must be set to full-duplex mode.  
ˆ You can create an aggregate trunk of transceivers with 1000Base-X  
fiber optic ports.  
ˆ Only those ports that are members of an aggregator transmit LACPDU  
packets.  
ˆ The load distribution method is applied at the aggregator level. To  
assign different load distribution methods to aggregate trunks, you  
must create a separate aggregator for each trunk. For further  
ˆ A member port of an aggregator functions as part of an aggregate  
trunk only if it receives LACPDU packets from the remote device. If it  
does not receive LACPDU packets, it functions as a regular Ethernet  
port, forwarding network traffic while also continuing to transmit  
LACPDU packets.  
ˆ The port with the highest priority in an aggregate trunk carries  
broadcast packets and packets with an unknown destination.  
90  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
ˆ When creating a new aggregator, you can specify either a name for the  
aggregator or an adminkey, but not both. If you specify a name, the  
adminkey is based on the operator key of the lowest numbered port in  
the aggregator. If you specify an adminkey, the default name is  
DEFAULT_AGG followed by the port number of the lowest numbered  
port in the aggregator. For example, an aggregator of ports 12 to 16 is  
assigned the default name DEFAULT_AGG12.  
ˆ Prior to creating an aggregate trunk between an Allied Telesis device  
and another vendor’s device, refer to the vendor’s documentation to  
determine the maximum number of active ports the device can support  
in a trunk. If the number is less than eight, the maximum number for  
the AT-9400 Switch, you should probably assign it a higher system  
LACP priority than the AT-9400 Switch. If it is more than eight, assign  
the AT-9400 Switch the higher priority. This can help avoid a possible  
conflict between the devices if some ports are placed in the standby  
mode when the devices create the trunk. For background information,  
ˆ LACPDU packets are transmitted as untagged packets.  
Section I: Basic Operations  
91  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 6: LACP Port Trunks  
92  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 7  
Port Mirror  
This chapter explains the port mirror feature. Sections in the chapter  
include:  
Section I: Basic Operations  
93  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 7: Port Mirror  
Supported Platforms  
This feature is supported on all AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
94  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
The port mirror feature allows for the unobtrusive monitoring of ingress or  
egress traffic on one or more ports on a switch, without impacting network  
performance or speed. It copies the traffic from specified ports to another  
switch port where the traffic can be monitored with a network analyzer.  
The port(s) whose traffic is mirrored is called the source port(s). The port  
where the traffic is copied to is referred to as the destination port.  
Guidelines  
Observe the following guidelines when creating a port mirror:  
ˆ A standalone switch can have only one destination port.  
ˆ A stack of Basic Layer 3 switches and the AT-StackXG Stacking  
Module can have only one destination port.  
ˆ You can mirror more than one source port at a time. However, the  
destination port may have to discard packets if the source ports are  
very active.  
ˆ In a stand-alone switch the source and destination ports must be  
located on the same switch.  
ˆ For a stack of Basic Layer 3 switches and the AT-StackXG Stacking  
Module, the destination and source ports of a port mirror can be  
located on different switches in the same stack.  
ˆ You can mirror the ingress or egress traffic of the source ports, or both.  
ˆ To create a mirror port for the Denial of Service defenses, specify only  
the destination port for the mirrored traffic. The management software  
automatically determines the source ports.  
Section I: Basic Operations  
95  
Download from Www.Somanuals.com. All Manuals Search And Download.  
           
Chapter 7: Port Mirror  
96  
Section I: Basic Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
98  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 8  
File System  
The chapter explains the switch’s file system and contains the following  
sections:  
Section II: Advanced Operations  
99  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 8: File System  
Overview  
The AT-9400 Switch has a file system in flash memory for storing system  
files. You can view a list of the files as well as copy, rename, and delete  
files. For those AT-9400 Switches that support a compact flash memory  
card, you can perform the same functions on the files stored on a flash  
card, as well as copy files between the switch’s file system and a flash  
card.  
The file system supports the following file types:  
ˆ Configuration files  
ˆ Public keys  
ˆ CA and self-signed certificates  
ˆ Certificate enrollment requests  
ˆ Event logs  
For an explanation of a boot configuration file, refer to “Boot Configuration  
Public encryption keys, public certificates, and certificate enrollment  
request files are related to the Secure Sockets Layer (SSL) certificates  
chapters for background information on those files.  
Note  
The certificate file, certificate enrollment request file, and key file are  
supported only on the version of AT-S63 Management Software that  
features SSL and PKI security.  
Note  
The file system may contain one or more ENC.UKF files. These are  
encryption key pairs. These files cannot be deleted, copied, or  
exported from the file system. For further information, refer to  
100  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Boot Configuration Files  
A boot configuration file contains the series of commands that recreate the  
current or a specific configuration of the switch when the unit is power  
cycled or reset. The commands in the file recreate all the VLANs, port  
settings, spanning tree settings, port trunks, port mirrors, and so forth.  
A switch can contain multiple boot configuration files, but only one can be  
active on a switch at a time. The active boot file is the file that is updated  
whenever you select the Save Configuration Changes option from the  
Main Menu.  
You can create different boot configuration files and store them in the  
switch’s file system. For example, you might create a backup of a boot  
configuration file to protect against the loss of the file, or you might create  
different boot configuration files to see which works best on the switch and  
for your network. You can also copy boot configuration files onto different  
switches to save yourself the trouble of having to manually configure  
AT-9400 Series switches that are to have similar configurations. One way  
to do this with switches that support compact flash cards is to copy the  
configuration file from flash memory on the master switch onto the  
compact flash card. Then take the compact flash card to other switches  
and copy the configuration file from the compact flash card into the  
switch’s flash memory.  
Section II: Advanced Operations  
101  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 8: File System  
File Naming Conventions  
The flash memory file system is a flat file system—directories are not  
supported. However, directories are supported on compact flash cards. In  
both types of storage, files are uniquely identified by a file name in the  
following format:  
filename.ext  
where:  
ˆ filename is a descriptive name for the file, and may be one to sixteen  
characters in length. Valid characters are lowercase letters (a–z),  
uppercase letters (A–Z), digits (0–9), and the following characters: ~ ’  
@ # $ % ^ & ( ) _ - { }. Invalid characters are: ! * + = “| \ [ ] ; : ? / , < >.  
ˆ ext is a file name extension of three characters in length, preceded by  
a period (.). The extension is used by the switch to determine the file  
type.  
Table 9. File Extensions and File Types  
Extension  
File Type  
Configuration file  
.cfg  
.cer  
.csr  
.key  
.log  
Certificate file  
Certificate enrollment request  
Public encryption key  
Event log  
The following is an example of a valid file name for a boot configuration  
file:  
standardconfig.cfg  
The following is an example of an invalid file name for a file stored in flash  
memory:  
sys/head_o.cfg  
The backslash character (/ ) is not a valid character for files stored in flash  
memory because subdirectories are not supported in the flash memory  
system.  
The file system displays filenames and directories in DOS 28.3 format.  
Filenames and directories longer than 32 bytes are represented in DOS  
8.3 format.  
102  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Using Wildcards to Specify Groups of Files  
You can use the asterisk character (*) as a wildcard character in some  
fields to identify groups of files. In addition, a wildcard can be combined  
with other characters. The following are examples of valid wildcard  
expressions:  
*.cfg  
*.key  
28*.cfg  
Section II: Advanced Operations  
103  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 8: File System  
104  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 9  
Event Logs and the Syslog Client  
This chapter describes how to monitor the activity of a switch by viewing  
the event messages in the event logs and sending the messages to a  
syslog server. Sections in the chapter include:  
Section II: Advanced Operations  
105  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 9: Event Logs and the Syslog Client  
Supported Platforms  
This feature is supported on all AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
106  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
A managed switch is a complex piece of computer equipment that includes  
both hardware and software. Multiple software features operate  
simultaneously, interoperating with each other and processing large  
amounts of network traffic. It is often difficult to determine exactly what is  
happening when a switch appears not to be operating normally, or what  
happened when a problem occurred.  
The operation of the switch can be monitored by viewing the event  
messages generated by the device. These events and the vital information  
about system activity that they provide can help identify and solve system  
problems.  
Event Messages  
Event messages include the following information:  
ˆ The time and date of the event  
ˆ The severity of the event  
ˆ The management module that generated the event  
ˆ An event description  
The switch has two event logs for storing the event messages. One log is  
located in temporary memory and has a storage capacity of up to 4,000  
entries. The events in this log are purged whenever you reset or power  
cycle the switch. The second log is located in permanent memory and has  
a maximum storage capacity of 2,000 entries. Events in this log are  
retained when the switch is reset or power cycled. Both logs store the  
same events messages. You can view either log to display the events of  
the switch since the unit was last reset. But to view the events that  
preceded a system reset, you must view the permanent event log.  
Section II: Advanced Operations  
107  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 9: Event Logs and the Syslog Client  
Syslog Client  
The management software features a syslog client for sending event  
messages to a syslog server on your network. A syslog server can  
function as a central repository for events from many different network  
devices.  
In order for the switch to send events to a syslog server, you must define a  
syslog output by specifying the IP address of the syslog server along with  
other information, such as the types of event messages the switch is to  
send to the server. You can create up to 19 syslog definitions on the  
switch.  
Note  
The event logs, even when disabled, log all AT-S63 initialization  
events that occur when the switch is reset or power cycled. Any  
switch events that occur after AT-S63 initialization are entered into  
the logs only if you enable the event log feature. The default setting  
for the event log feature is enabled.  
Observe the following guidelines when using this feature:  
ˆ You can define up to 19 log output definitions.  
ˆ The event logs on the switch must be activated in order for the switch  
to send events to a syslog server.  
ˆ The local subnet on the switch where the syslog server is a member  
must have a routing interface. The switch uses the IP address of the  
routing interface as its source address when communicating with the  
server. For background information on routing interfaces, refer to  
Note  
Prior to version 2.0.0 of the AT-S63 management software, a syslog  
server had to be a member of the switch’s management VLAN. This  
restriction no longer applies. The server can be located on any local  
subnet of the switch, so long as the subnet has been assigned a  
routing interface on the switch.  
108  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 10  
Classifiers  
This chapter explains classifiers for access control lists and Quality of  
Service policies. The sections in this chapter include:  
Section II: Advanced Operations  
109  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 10: Classifiers  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three of the management interfaces  
in the AT-S63 Management Software:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
110  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Overview  
A classifier defines a traffic flow. A traffic flow consists of packets that  
share one or more characteristics. A traffic flow can range from being very  
broad to very specific. An example of the former might be all IP traffic while  
an example of the latter could be packets with specific source and  
destination MAC addresses.  
A classifier contains a set of criteria for defining a traffic flow. Examples of  
the variables include source and destination MAC addresses, source and  
destination IP addresses, IP protocols, source and destination TCP and  
UDP ports numbers, and so on. You can also specify more than one  
criteria within a classifier to make the definition of the traffic flow more  
specific. Some of the variables you can mix-and-match, but there are  
restrictions, as explained later in this section in the descriptions of the  
individual variables.  
By itself, a classifier does not perform any action or produce any result  
because it lacks instructions on what a port should do when it receives a  
packet that belongs to the defined traffic flow. Rather, the action is  
established outside the classifier. As a result, you will never use a  
classifier by itself.  
There are two AT-S63 features that use classifiers. They are:  
ˆ Access control lists (ACL)  
ˆ Quality of Service (QoS) policies  
filters ingress packets on a port by controlling which packets a port will  
accept and reject. You can use this feature to improve the security of your  
network or enhance network performance by creating network paths or  
links dedicated to carrying specific types of traffic.  
When you create an ACL you must specify the traffic flow you want the  
ACL to control. You do that by creating one or more classifiers and adding  
the classifiers to the ACL. The action that the port takes when an ingress  
packet matches the traffic flow specified by a classifier is contained in the  
ACL itself. The action will be to either accept packets of the traffic flow or  
discard them.  
The other feature that uses classifiers is Quality of Service (QoS) policies.  
You can use this feature to regulate the various traffic flows that pass  
through the switch. For instance, you might raise or lower their user priority  
values or increase or decrease their allotted bandwidths.  
As with an ACL, you specify the traffic flow of interest by creating one or  
more classifiers and applying them to a QoS policy. The action to be taken  
by a port when it receives a packet that corresponds to the prescribed flow  
Section II: Advanced Operations  
111  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 10: Classifiers  
is dictated by the QoS policy, as explained in Chapter 13, “Quality of  
In summary, a classifier is a list of variables that define a traffic flow. You  
apply a classifier to an ACL or a QoS policy to define the traffic flow you  
want the ACL or QoS policy to affect or control.  
112  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Classifier Criteria  
The components of a classifier are defined in the following subsections.  
Destination MAC Address (Layer 2)  
Source MAC Address (Layer 2)  
You can identify a traffic flow by specifying a source and/or destination  
MAC address. For instance, you might create a classifier for a traffic flow  
destined to a particular destination node, or from a specific source node to  
a specific destination node, all identified by their MAC addresses.  
The management software does not support a classifier based on a range  
of MAC addresses. Different MAC addresses must be considered  
separate traffic flows, with their own classifiers.  
Ethernet 802.2 and Ethernet II Frame Types (Layer 2)  
You can create a classifier that filters packets based on Ethernet frame  
type and whether a packet is tagged or untagged within a frame type. (A  
tagged Ethernet frame contains within it a field that specifies the ID  
number of the VLAN to which the frame belongs. Untagged packets lack  
this field.) Options are:  
ˆ Ethernet II tagged packets  
ˆ Ethernet II untagged packets  
ˆ Ethernet 802.2 tagged packets  
ˆ Ethernet 802.2 untagged packets  
802.1p Priority Level (Layer 2)  
page 257, contains within it a field that specifies its VLAN membership.  
Such frames also contain a user priority level used by the switch to  
determine the Quality of Service to apply to the frame and which egress  
queue on the egress port a packet should be stored in. The three bit binary  
number represents eight priority levels, 0 to 7, with 0 the lowest priority  
and 7 the highest. Figure 4 illustrates the location of the user priority field  
within an Ethernet frame.  
Section II: Advanced Operations  
113  
Download from Www.Somanuals.com. All Manuals Search And Download.  
           
Chapter 10: Classifiers  
Destination  
Address  
Source  
Address  
Type/  
Length  
Preamble  
64 bits  
Frame Data  
CRC  
16  
bits  
48 bits  
48 bits  
368 to 12000 bits  
32 bits  
User  
Priority  
Tag Protocol Identifier  
16 bits  
CFI  
VLAN Identifier  
12 bits  
1
bit  
3 bits  
Figure 4. User Priority and VLAN Fields within an Ethernet Frame  
You can identify a traffic flow of tagged packets using the user priority  
value. A classifier for such a traffic flow would instruct a port to watch for  
tagged packets containing the specified user priority level.  
The priority level criteria can contain only one value, and the value must  
be from 0 (zero) to 7. Multiple classifiers are required if a port is to watch  
for several different traffic flows of different priority levels.  
VLAN ID (Layer 2)  
A tagged Ethernet frame also contains within it a field of 12 bits that  
specifies the ID number of the VLAN to which the frame belongs. The  
field, illustrated in Figure 4, can be used to identify a traffic flow.  
A classifier can contain only one VLAN ID. To create a port ACL or QoS  
policy that applies to several different VLAN IDs, multiple classifiers are  
required.  
Protocol (Layer 2)  
Traffic flows can be identified by the protocol specified in the Ethertype  
field of the MAC header in an Ethernet II frame. Possible values are:  
ˆ IP  
ˆ ARP  
ˆ RARP  
ˆ Protocol Number  
114  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Observe the following guidelines when using this variable:  
ˆ When selecting a Layer 3 or Layer 4 variable, this variable must be left  
blank or set to IP.  
ˆ If you choose to specify a protocol by its number, you can enter the  
value in decimal or hexadecimal format. If you choose the latter,  
precede the number with the prefix “0x”.  
ˆ The range for the protocol number is 1536 (0x600) to 65535 (0xFFFF).  
IP ToS (Type of Service) (Layer 3)  
Type of Service (ToS) is a standard field in IP packets. It is used by  
applications to indicate the priority and Quality of Service for a frame. The  
range of the value is 0 to 7. The location of the field is shown in Figure 5.  
0
3
7
31  
15  
ver IHL  
ToS  
total length  
. . .  
0
1
2
3
4
5
6
7
DSCP value  
Precedence  
Figure 5. ToS field in an IP Header  
Observe these guidelines when using this criterion:  
ˆ The Protocol variable must be left blank or set to IP.  
ˆ You cannot specify both an IP ToS value and an IP DSCP value in the  
same classifier.  
IP DSCP (DiffServ Code Point) (ToS) (Layer 3)  
The Differentiated Services Code Point (DSCP) tag indicates the class of  
service to which packets belong. The DSCP value is written into the TOS  
field of the IP header, as shown in Figure 5 on page 115. Routers within  
the network use this DSCP value to classify packets, and assign QoS  
appropriately. When a packet leaves the DiffServ domain, the DSCP value  
can be replaced with a value appropriate for the next DiffServ domain.The  
range of the value is 0 to 63.  
Section II: Advanced Operations  
115  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 10: Classifiers  
Observe these guidelines when using this criterion:  
ˆ The Protocol variable must be left blank or set to IP.  
ˆ You cannot specify both an IP ToS value and an IP DSCP value in the  
same classifier.  
IP Protocol (Layer 3)  
You can define a traffic flow by the following Layer 3 protocols:  
ˆ TCP  
ˆ UDP  
ˆ ICMP  
ˆ IGMP  
ˆ IP protocol number  
If you choose to specify the protocol by its number, you can enter the  
value in decimal or hexadecimal format. It the latter, include the prefix “0x”.  
The range for the protocol number is 0 (0x0) to 255 (0xFF).  
Source IP Addresses (Layer 3)  
Source IP Mask (Layer 3)  
You can define a traffic flow by the source IP address contained in IP  
packets. The address can be of a subnet or a specific end node.  
You do not need to enter a source IP mask if you are filtering on the IP  
address of a specific end node. A mask is required, however, when you  
filter on a subnet. A binary “1” indicates the switch should filter on the  
corresponding bit of the IP address, while a “0” indicates that it should not.  
For example, the subnet address 149.11.11.0 would have the mask  
“255.255.255.0.”  
Observe this guideline when using these criteria:  
ˆ The Protocol variable must be left blank or set to IP.  
Destination IP Addresses (Layer 3)  
Destination IP Mask (Layer 3)  
You can also define a traffic flow based on the destination IP address of a  
subnet or a specific end node.  
You do not need to enter a destination IP mask for an IP address of a  
specific end node. A mask is required, however, when filtering on a  
subnet. A binary “1” indicates the switch should filter on the corresponding  
bit of the IP address while a “0” indicates that it should not. For example,  
the subnet address 149.11.11.0 would have the mask “255.255.255.0.”  
116  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Observe this guideline when using these criteria:  
ˆ The Protocol variable must be left blank or set to IP.  
TCP Source Ports (Layer 4)  
TCP Destination Ports (Layer 4)  
A traffic flow can be identified by a source and/or destination TCP port  
number contained within the header of an IP frame. Observe the following  
guidelines when using these criteria:  
ˆ The Protocol variable must be left blank or set to IP.  
ˆ The IP Protocol variable must be left blank or set to TCP.  
ˆ A classifier cannot contain criteria for both TCP and UDP ports. You  
may specify one in a classifier, but not both.  
UDP Source Ports (Layer 4)  
UDP Destination Ports (Layer 4)  
A traffic flow can be identified by a source and/or destination UDP port  
number contained within the header of an IP frame. Observe the following  
guidelines when using these criteria:  
ˆ The Protocol variable must be left blank or set to IP.  
ˆ The IP Protocol variable must be left blank or set to UDP.  
ˆ A classifier cannot contain criteria for both TCP and UDP ports. You  
may specify only one in a classifier.  
TCP Flags  
A traffic flow can be based on the following TCP flags:  
ˆ URG - Urgent  
ˆ ACK - Acknowledgement  
ˆ RST - Reset  
ˆ PSH - Push  
ˆ SYN - Synchronization  
ˆ FIN - Finish  
Observe the following guidelines when using this criterion:  
ˆ The Protocol variable must be left blank or set to IP.  
ˆ The IP Protocol variable must be left blank or set to TCP.  
ˆ A classifier cannot contain both a TCP flag and a UDP source and/or  
destination port.  
Section II: Advanced Operations  
117  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 10: Classifiers  
Guidelines  
Follow these guidelines when creating a classifier:  
ˆ Each classifier represents a separate traffic flow.  
ˆ The variables within a classifier are linked by AND. The more variables  
defined within a classifier, the more specific it becomes in terms of the  
flow it defines. For instance, specifying both a source IP address and a  
TCP destination port within the same classifier defines a traffic flow  
that relates to IP packets containing both the designated source IP  
address and TCP destination port. However, there are some  
restrictions on combining variables in the same classifier. For the  
ˆ The same classifier can be applied to both ACLs and QoS policies.  
ˆ You can apply the same classifier to more than one ACL or QoS  
policy.  
ˆ A classifier without any defined variables applies to all packets.  
ˆ You cannot create two classifiers that have the same settings. There  
can be only one classifier for any given type of traffic flow.  
ˆ A classifier can have a maximum of eight defined criteria, not including  
the classifier ID number and the description.  
ˆ The switch can store up to 256 classifiers. However, the maximum  
number of classifiers that you can assign to active access control lists  
and QoS policies at any one time will be from 14 to 127. The number  
depends on several factors, such as the number of ports to which the  
classifiers are assigned and the types of criteria defined in the  
classifiers.  
ˆ You cannot modify a classifier if it belongs to an ACL or QoS policy  
assigned to a port. You must remove the port assignments from the  
ACL or policy and reassign them after modifying the classifier.  
ˆ You cannot delete a classifier if it belongs to an ACL or QoS policy.  
You must remove a classifier from its ACLs and QoS policies before  
you can delete it.  
118  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 11  
Access Control Lists  
This chapter describes access control lists (ACL) and how they can  
improve network security and performance. This chapter contains the  
following sections:  
Section II: Advanced Operations  
119  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 11: Access Control Lists  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Switches  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Switches  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three management interfaces in the  
AT-S63 Management Software:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
120  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Overview  
An access control list is a filter that controls the ingress traffic on a port. It  
defines a category of traffic and the action of the port when it receives  
packets of the category. The action can be to accept the defined packets  
or discard them. You can use this feature to increase the security to your  
network by restricting access to certain areas or subnets, or to enhance  
network performance by forming network links dedicated to carrying  
specified types of traffic.  
Note  
This feature is not related to the management ACL feature,  
page 431. They perform different functions and are configured in  
different ways.  
The heart of an ACL is a classifier. A classifier, as explained “Overview” on  
page 111, defines packets that share a common trait. Packets that share a  
trait are referred to as a traffic flow. A traffic flow can be very broad, such  
as all IP packets, or very specific, such as packets from a specified end  
node destined for another specified node. You specify the traffic using  
different criteria, such as source and destination MAC addresses or  
protocol.  
When you create an ACL, you must specify the classifier that defines the  
traffic flow to permit or deny on a port.  
There are two kinds of ACLs based on the two actions that an ACL can  
perform. One is called a permit ACL. Packets that meet the criteria in a  
permit ACL are accepted by a port.  
The second type of ACL is a deny ACL. This type of ACL denies entry to  
packets that meet the criteria of its classifiers, unless the packet also  
meets the criteria of a permit ACL on the same port, in which case the  
packet is accepted. This is because a permit ACL overrides a deny ACL.  
Here is an overview of how the process works.  
1. When an ingress packet arrives on a port, it is checked against the  
criteria in the classifiers of all the ACLs, both permit and deny,  
assigned to the port.  
2. If the packet matches the criteria of a permit ACL, the port immediately  
accepts it, even if the packet also matches a deny ACL assigned to the  
same port, because a permit ACL always overrides a deny ACL.  
3. If a packet meets the criteria of a deny ACL but not any permit ACLs  
on the port, then the packet is discarded.  
Section II: Advanced Operations  
121  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 11: Access Control Lists  
4. Finally, if a packet does not meet the criteria of any ACLs on a port, it  
is accepted by the port.  
122  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Parts of an ACL  
An ACL must have the following information:  
ˆ Name - An ACL must have a name. The name of an ACL should  
indicate the type of traffic flow being filtered and, perhaps, also the  
action. An example might be “HTTPS flow - permit.” The more specific  
the name, the easier it will be for you to identify it.  
ˆ Action - The action of an ACL can be permit or deny. Ingress traffic that  
meets the criteria of an ACL with the permit action is accepted by a  
port. Ingress traffic that meets the criteria of an ACL with the deny  
action is discarded by a port, unless the traffic also meets the criteria of  
a permit ACL on the same port, in which case it is accepted.  
ˆ Classifiers - An ACL must have at least one classifier. An ACL can  
have more than one classifier to filter multiple traffic flows.  
ˆ Port Lists - Finally, you need to specify the ports for the ACL.  
Section II: Advanced Operations  
123  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 11: Access Control Lists  
Guidelines  
Here are the rules to creating ACLs:  
ˆ A port can have multiple permit and deny ACLs.  
ˆ An ACL must have at least one classifier.  
ˆ An ACL can be assigned to more than one switch port.  
ˆ An ACL filters ingress traffic, but not egress traffic.  
ˆ The action of a ACL can be either permit or deny. A permit ACL  
overrides a deny ACL on the same port when the ACLs define the  
same traffic.  
ˆ The order in which the ACLs are added to a port is not important since  
the packets are compared against all of a port’s ACLs.  
ˆ Since classifiers cannot be assigned more than once to a port, ACLs  
that have the same classifier cannot be assigned to the same port.  
ˆ An ACL and a Quality of Service policy cannot be assigned to the  
same port if they have a common classifier.  
ˆ The switch can store up to 64 ACLs.  
124  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Examples  
This section contains several examples of ACLs.  
In this example, port 4 has been assigned one ACL, a deny ACL for the  
subnet 149.11.11.0. This ACL prevents the port from accepting any traffic  
originating from that subnet. Since this is the only ACL on the port, all  
other traffic is accepted. As explained earlier, a port automatically accepts  
all packets that do not meet the criteria of the classifiers assigned to its  
ACLs.  
Create Classifier  
Create Access Control Lists (ACL)  
1 - ACL ID ................. 4  
01 - Classifier ID: ..... 22  
2 - Description .......... 149.11.11-deny  
3 - Action .................. Deny  
4 - Classifier List ...... 22  
02 - Description: ...... 149.11.11 flow  
.
.
5 - Port List .............. 4  
12 - Src IP Addr: ..... 149.11.11.0  
13 - Src IP Mask ..... 255.255.255.0  
Figure 6. ACL Example 1  
Section II: Advanced Operations  
125  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 11: Access Control Lists  
To deny traffic from several subnets on the same port, you can create  
multiple classifiers and apply them to the same ACL, as illustrated in the  
next example. Three subnets are denied access to port 4. The three  
classifiers defining the subnets are applied to the same ACL.  
Create Classifier  
01 - Classifier ID: ..... 22  
02 - Description: ...... 149.11.11 flow  
.
.
12 - Src IP Addr: ..... 149.11.11.0  
13 - Src IP Mask: .... 255.255.255.0  
Create Classifier  
Create Access Control Lists (ACL)  
01 - Classifier ID: ..... 24  
02 - Description: ...... 149.22.22 flow  
.
.
12 - Src IP Addr: ..... 149.22.22.0  
13 - Src IP Mask: .... 255.255.255.0  
1 - ACL ID ................. 4  
2 - Description .......... Subnets - deny  
3 - Action .................. Deny  
4 - Classifier List ...... 22, 24, 62  
5 - Port List .............. 4  
Create Classifier  
01 - Classifier ID: ..... 62  
02 - Description: ...... 149.33.33 flow  
.
.
12 - Src IP Addr: ..... 149.33.33.0  
13 - Src IP Mask: .... 255.255.255.0  
Figure 7. ACL Example 2  
126  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
The same result can be achieved by assigning the classifiers to different  
ACLs and assigning the ACLs to the same port, as in this example, again  
for port 4.  
Create Classifier  
Create Access Control Lists (ACL)  
01 - Classifier ID: ..... 22  
1 - ACL ID ................. 4  
02 - Description: ...... 149.11.11 flow  
.
.
2 - Description .......... 149.11.11-deny  
3 - Action .................. Deny  
4 - Classifier List ...... 22  
12 - Src IP Addr: ..... 149.11.11.0  
13 - Src IP Mask: .... 255.255.255.0  
5 - Port List .............. 4  
Create Classifier  
Create Access Control Lists (ACL)  
01 - Classifier ID: ..... 24  
1 - ACL ID ................. 22  
02 - Description: ...... 149.22.22 flow  
.
.
2 - Description .......... 149.22.22.-deny  
3 - Action .................. Deny  
4 - Classifier List ...... 24  
12 - Src IP Addr: ..... 149.22.22.0  
13 - Src IP Mask: .... 255.255.255.0  
5 - Port List .............. 4  
Create Classifier  
Create Access Control Lists (ACL)  
1 - ACL ID ................. 23  
01 - Classifier ID: ..... 62  
2 - Description .......... 149.33.33-deny  
3 - Action .................. Deny  
4 - Classifier List ...... 62  
02 - Description: ...... 149.33.33 flow  
.
.
12 - Src IP Addr: ..... 149.33.33.0  
13 - Src IP Mask: .... 255.255.255.0  
5 - Port List .............. 4  
Figure 8. ACL Example 3  
Section II: Advanced Operations  
127  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 11: Access Control Lists  
In this example, the traffic on ports 14 and 15 is restricted to packets from  
the source subnet 149.44.44.0. All other IP traffic is denied. Classifier ID  
11, which specifies the traffic flow to be permitted by the ports, is assigned  
to an ACL with an action of permit. Classifier ID 17 specifies all IP traffic  
and is assigned to an ACL whose action is deny. Since a permit ACL  
overrides a deny ACL, the port will accept the traffic from the 149.44.44.0  
subnet, while discarding all other IP traffic, even though that traffic also  
happens to meet the criteria of the deny ACL.  
Create Access Control Lists (ACL)  
Create Classifier  
01 - Classifier ID: ..... 11  
1 - ACL ID ................. 21  
2 - Description .......... 149.44.44-permit  
3 - Action .................. Permit  
4 - Classifier List ...... 11  
02 - Description: ....... 149.44.44-flow  
.
.
12 - Src IP Addr: ....... 149.44.44.0  
13 - Src IP Mask: ...... 255.255.255.0  
5 - Port List .............. 14,15  
Create Access Control Lists (ACL)  
Create Classifier  
1 - ACL ID ................. 5  
01 - Classifier ID: ..... 17  
02 - Description: ....... All IP flow  
.
.
2 - Description .......... All IP - deny  
3 - Action .................. Deny  
4 - Classifier List ...... 17  
5 - Port List .............. 14,15  
08 - Protocol: ............ IP  
Figure 9. ACL Example 4  
This example limits the traffic on port 22 to HTTPS web traffic intended for  
the end node with the IP address 149.55.55.55, and rejects all other IP  
traffic. (The Dst IP Mask field in classifier 6 is left empty because a mask is  
not required for a source or destination IP address for a specific end node.  
If you wanted to include it, it would be 255.255.255.255.)  
Create Classifier  
Create Access Control Lists (ACL)  
01 - Classifier ID: ...... 6  
02 - Description: ....... 55.55 HTTPS  
.
.
14 - Dst IP Addr: ....... 149.55.55.55  
15 - Dst IP Mask: ......  
.
1 - ACL ID ................. 4  
2 - Description .......... Web - permit  
3 - Action .................. Permit  
4 - Classifier List ...... 6  
5 - Port List .............. 22  
17 - TCP Dst Port: ..... 443  
Create Access Control Lists (ACL)  
Create Classifier  
1 - ACL ID ................. 5  
01 - Classifier ID: ..... 17  
02 - Description: ....... All IP flow  
.
.
2 - Description .......... All IP - deny  
3 - Action .................. Deny  
4 - Classifier List ...... 17  
5 - Port List .............. 22  
08 - Protocol: ............ IP  
Figure 10. ACL Example 5  
128  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
The next example limits the ingress traffic on port 17 to IP packets from  
the subnet 149.22.11.0 and a Type of Service setting of 6, destined to the  
end node with the IP address 149.22.22.22. All other IP traffic and ARP  
packets are prohibited.  
Create Classifier  
Create Access Control Lists (ACL)  
01 - Classifier ID: ...... 6  
02 - Description: ....... ToS 6 subnet flow  
.
.
09 - IP ToS: ............... 6  
.
1 - ACL ID ................. 4  
2 - Description .......... ToS 6 traffic - permit  
3 - Action .................. Permit  
4 - Classifier List ...... 6  
5 - Port List .............. 17  
12 - Src IP Addr: ....... 149.22.11.0  
13 - Src IP Mask: ...... 255.255.255.0  
14 - Dst IP Addr: ....... 149.22.22.22  
15 - Dst IP Mask: ......  
Create Access Control Lists (ACL)  
Create Classifier  
1 - ACL ID ................. 23  
01 - Classifier ID: ..... 8  
02 - Description: ...... All IP flow  
.
.
2 - Description .......... All IP flow - deny  
3 - Action .................. Deny  
4 - Classifier List ...... 8,67  
5 - Port List .............. 17  
08 - Protocol: ........... IP  
Create Classifier  
01 - Classifier ID: ..... 67  
02 - Description: ...... All ARP flow  
.
.
08 - Protocol: ........... 0x806 (ARP)  
Figure 11. ACL Example 6  
Section II: Advanced Operations  
129  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 11: Access Control Lists  
130  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 12  
Class of Service  
This chapter describes the Class of Service (CoS) feature. Sections in the  
chapter include:  
Section II: Advanced Operations  
131  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 12: Class of Service  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
132  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
When a port on an Ethernet switch becomes oversubscribed—its egress  
queues contain more packets than the port can handle in a timely  
manner—the port may be forced to delay the transmission of some  
packets, resulting in the delay of packets reaching their destinations. A  
port may be forced to delay transmission of packets while it handles other  
traffic. Some packets destined to be forwarded to an oversubscribed port  
from other switch ports may be discarded.  
Although minor delays are often of no consequence to a network or its  
performance, there are applications, referred to as delay or time sensitive  
applications, that can be impacted by packet delays. Voice transmission  
and video conferencing are two examples. A delay in the transmission of  
packets carrying their data could impact the quality of the audio or video.  
This is where CoS can be of value. What it does is it permits a switch to  
give higher priority to some packets over other packets.  
There are two principal types of traffic found on the ports of a Gigabit  
Ethernet switch, one being untagged packets and the other tagged  
packets. As explained in “Tagged VLAN Overview” on page 257, one of  
the principal differences between them is that tagged packets contain  
VLAN information.  
CoS applies mainly to tagged packets because, in addition to carrying  
VLAN information, these packets can also contain a priority level  
specifying how important (delay sensitive) a packet is in comparison to  
other packets. It is this number that the switch refers to when determining  
a packet’s priority level.  
CoS, as defined in the IEEE 802.1p standard, has eight levels of priority.  
The priorities are 0 to 7, with 0 the lowest priority and 7 the highest.  
Each switch port has eight egress queues, labeled Q0, Q1, Q2, Q3, Q4,  
Q5, Q6, Q7. Q0 is the lowest priority queue and Q7 is the highest. A  
packet in a high priority egress queue is typically transmitted out a port  
sooner than a packet in a low priority queue.  
When a tagged packet arrives on a port, the switch examines its priority  
value to determine which egress priority queue the packet should be  
directed to on the egress port. Table 10 lists the default mappings between  
the eight CoS priority levels and the eight egress queues of a switch port.  
Section II: Advanced Operations  
133  
Download from Www.Somanuals.com. All Manuals Search And Download.  
         
Chapter 12: Class of Service  
Table 10. Default Mappings of IEEE 802.1p Priority Levels to Priority  
Queues  
IEEE 802.1p Priority  
Port Priority Queue  
Level  
0
1
2
3
4
5
6
7
Q1  
Q0 (lowest)  
Q2  
Q3  
Q4  
Q5  
Q6  
Q7 (highest)  
For example, when a tagged packet with a priority level of 3 enters a port  
on the switch, the packet is stored in Q3 queue on the egress port.  
Note that priority 0 is mapped to CoS queue 1 instead of CoS queue 0  
because tagged traffic that has never been prioritized has a VLAN tag  
User Priority of 0. If priority 0 was mapped to CoS queue 0, this default  
traffic goes to the lowest queue, which is probably undesirable. This  
mapping also makes it possible to give some traffic a lower priority than  
the default traffic.  
You can change these mappings. For example, you might decide that  
packets with a priority of 5 should be handled by egress queue Q3 and  
packets with a priority of 2 should be handled in Q1. The result is shown in  
Table 11. Customized Mappings of IEEE 802.1p Priority Levels to Priority  
Queues  
IEEE 802.1p Priority  
Port Priority Queue  
Level  
0
1
2
3
4
5
Q1  
Q0 (lowest)  
Q1  
Q3  
Q4  
Q3  
134  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Table 11. Customized Mappings of IEEE 802.1p Priority Levels to Priority  
Queues (Continued)  
IEEE 802.1p Priority  
Port Priority Queue  
Level  
6
7
Q6  
Q7 (highest)  
Note that because all ports must use the same priority-to-egress queue  
mappings, these mappings are applied at the switch level. They cannot be  
set on a per-port basis.  
You can configure a port to completely ignore the priority levels in its  
tagged packets and instead use a temporary priority level assigned to the  
port. For instance, perhaps you decide that all tagged packets received on  
port 4 should be assigned a priority level of 5, regardless of the priority  
level in the packets themselves.  
CoS relates primarily to tagged packets rather than untagged packets  
because untagged packets do not contain a priority level. By default, all  
untagged packets are assigned a priority of 0 and are placed in a port’s Q1  
egress queue. But you can override this and instruct a port’s untagged  
frames to be stored in a different priority queue.  
One last thing to note is that CoS does not change the priority level in a  
tagged packet. The packet leaves the switch with the same priority it had  
when it entered. This is true even if you change the default priority-to-  
egress queue mappings.  
Section II: Advanced Operations  
135  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 12: Class of Service  
Scheduling  
A switch port needs a mechanism for knowing the order in which it should  
handle the packets in its eight egress queues. For example, if all the  
queues contain packets, should the port transmit all packets from Q7, the  
highest priority queue, before moving on to the other queues, or should it  
instead just do a few packets from each queue and, if so, how many?  
This control mechanism is called scheduling. Scheduling determines the  
order in which a port handles the packets in its egress queues. The  
AT-S63 software has two types of scheduling:  
ˆ Strict priority  
ˆ Weighted round robin priority  
Note  
Scheduling is set at the switch level. You cannot set this on a per-  
port basis.  
Strict Priority With this type of scheduling, a port transmits all packets out of higher  
priority queues before transmitting any from the lower priority queues. For  
instance, as long as there are packets in Q7 it does not handle any  
packets in Q6.  
Scheduling  
The value to this type of scheduling is that high priority packets are always  
handled before low priority packets.  
The problem with this method is that some low priority packets might  
never be transmitted out the port because a port might never get to the low  
priority queues. A port handling a large volume of high priority traffic may  
be so busy transmitting that traffic that it never has an opportunity to get to  
any of the packets stored in its low priority queues.  
Weighted Round The weighted round robin scheduling method functions as its name  
implies. The port transmits a set number of packets from each queue, in a  
round robin fashion, so that each has a chance to transmit traffic. This  
method guarantees that every queue receives some attention from the  
Robin Priority  
Scheduling  
port for transmitting packets.  
To use this scheduling method, you need to specify the maximum number  
of packets a port should transmit from a queue before moving to the next  
queue. This is referred to as specifying the “weight” of a queue. In most  
cases, you will want to give greater weight to the higher priority queues  
over the lower priority queues.  
136  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
           
AT-S63 Management Software Features Guide  
Table 12 shows an example.  
Table 12. Example of Weighted Round Robin Priority  
Maximum Number of  
Packets  
Port Egress Queue  
Q0 (lowest)  
1
Q1  
Q2  
Q3  
Q4  
Q5  
Q6  
Q7  
1
5
5
5
5
10  
15  
In this example, the port transmits a maximum number of 15 packets from  
Q7 before moving to Q6, from where it transmits up to 10 packets, and so  
forth.  
For Q0 to Q6, the range of the maximum number of transmitted packets is  
1 to 15. The range for Q7, the highest priority queue, is 0 to 15. Setting Q7  
to 0 means its packets always take priority over the packets in the other  
queues and that no packets are transmitted from the lower priority queues  
so long as there are packets in Q7. This allows you to combine the two  
priority scheduling methods on the same port.  
An example.of Q7 with a weight of 0 is shown in Table 13. At these  
settings, a port transmits all of the packets from Q7 until the queue is  
empty, and then transmits a maximum of 15 packets from Q6, 8 packets  
from Q5, and so forth.  
Table 13. Example of a Weight of Zero for Priority Queue 7  
Maximum Number of  
Port Egress Queue  
Packets  
Q0 (lowest)  
1
1
8
8
8
8
Q1  
Q2  
Q3  
Q4  
Q5  
Section II: Advanced Operations  
137  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 12: Class of Service  
Table 13. Example of a Weight of Zero for Priority Queue 7 (Continued)  
Maximum Number of  
Port Egress Queue  
Packets  
Q6  
Q7  
15  
0
138  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 13: Quality of Service  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
140  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
Quality of Service allows you to prioritize traffic and/or limit the bandwidth  
available to it. The concept of QoS is a departure from the original  
networking protocols, which treated all traffic on the Internet or within a  
LAN in the same manner. Without QoS, every traffic type is equally likely  
to be dropped if a link becomes oversubscribed. This approach is now  
inadequate in many networks, because traffic levels have increased and  
networks transport time-critical applications such as streams of video and  
data. QoS also enables service providers to easily supply different  
customers with different amounts of bandwidth.  
Configuring Quality of Service involves two separate stages:  
ˆ Classifying traffic into flows, according to a wide range of criteria.  
Classification is performed by the switch’s packet classifiers, described  
ˆ Acting on these traffic flows.  
Quality of Service is a broadly used term that encompasses as a minimum  
both Layer 2 and Layer 3 in the OSI model. QoS is typically demonstrated  
by how the switch accomplishes the following:  
ˆ Assigns priority to incoming frames, if they do not carry priority  
information  
ˆ Maps prioritized frames to traffic classes, or maps frames to traffic  
classes based upon other criteria  
ˆ Maps traffic classes to egress queues, or maps prioritized frames to  
egress queues  
ˆ Provides maximum bandwidth limiting for traffic classes, egress  
queues and/or ports  
ˆ Schedules frames in egress queues for transmission (for example,  
empty queues in strict priority or samples each queue)  
ˆ Relabels the priority of frames  
ˆ Determines which frames to drop if the network becomes congested  
ˆ Reserves memory for switching/routing or QoS operation (e.g.  
reserving buffers for egress queues, or buffers to store packets with  
particular characteristics)  
Note  
QoS is only performed on packets that are switched at wire speed.  
This includes IP, IP multicast, IPX, and Layer 2 traffic within VLANs.  
Section II: Advanced Operations  
141  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 13: Quality of Service  
The QoS functionality described in this chapter sorts packets into various  
flows, according to the QoS policy that applies to the port the traffic is  
received on. The switch then allocates resources to direct this traffic  
according to bandwidth or priority settings in the policy. A policy contains  
traffic classes, flow groups, and classifiers. Therefore, to configure QoS,  
you:  
ˆ Create classifiers to sort packets into traffic flows.  
ˆ Create flow groups and add classifiers to them. Flow groups are  
groups of classifiers which group together similar traffic flows. You can  
apply QoS prioritization to flow groups.  
ˆ Create traffic classes and add flow groups to them. Traffic classes are  
groups of flow groups and are central to QoS. You can apply  
bandwidth limits and QoS prioritization to traffic classes.  
ˆ Create policies and add traffic classes to them. Policies are groups of  
traffic classes. A policy defines a complete QoS solution for a port or  
group of ports.  
ˆ Associate policies with ports.  
Note  
The steps listed above are in a conceptually logical order, but the  
switch cannot check a policy for errors until the policy is attached to  
a port. To simplify error diagnosis, define your QoS configuration on  
paper first, and then enter it into the management software starting  
with classifiers.  
Policies, traffic classes, and flow groups are created as individual entities.  
When a traffic class is added to a policy, a logical link is created between  
the two entities. Destroying the policy unlinks the traffic class, leaving the  
traffic class in an unassigned state. Destroying a policy does not destroy  
any of the underlying entities. Similarly, destroying a traffic class unlinks  
flow groups, and destroying flow groups unlinks classifiers.  
142  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Classifiers  
Classifiers identify a particular traffic flow, and range from general to  
Note that a single classifier should not be used in different flows that will  
end up, through traffic classes, assigned to the same policy. A classifier  
should only be used once per policy. Traffic is matched in the order of  
classifiers. For example, if a flow group has classifiers 1, 3, 2 and 5, that is  
the order in which the packets are matched.  
Section II: Advanced Operations  
143  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 13: Quality of Service  
Flow Groups  
Flow groups group similar traffic flows together, and allow more specific  
QoS controls to be used, in preference to those specified by the traffic  
class. Flow groups consist of a small set of QoS parameters and a group  
of classifiers. After a flow group has been added to a traffic class it cannot  
be added to another traffic class. A traffic class may have many flow  
groups. Traffic is matched in the order of the flow groups. For example, if a  
traffic class has flow groups 1, 3, 2 and 5, this is the order in which the  
packets are matched.  
QoS controls at the flow group level provide a QoS hierarchy. Non-default  
flow group settings are always used, but if no setting is specified for a flow  
group, the flow group uses the settings for the traffic class to which it  
belongs. For example, you can use a traffic class to limit the bandwidth  
available to web and FTP traffic combined. Within that traffic class, you  
can create two different flow groups with different priorities, to give web  
traffic a higher priority than FTP. Web traffic would then be given  
preferential access to bandwidth, but would be limited to the bandwidth  
limit of the traffic class.  
144  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Traffic Classes  
Traffic classes are the central component of the QoS solution. They  
provide most of the QoS controls that allow a QoS solution to be deployed.  
A traffic class can be assigned to only one policy. Traffic classes consist of  
a set of QoS parameters and a group of QoS flow groups. Traffic can be  
prioritized, marked (IP TOS or DSCP field set), and bandwidth limited.  
Traffic is matched in the order of traffic class. For example, if a policy has  
traffic classes 1, 3, 2 and 5, this is the order in which the packets are  
matched.  
Section II: Advanced Operations  
145  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 13: Quality of Service  
Policies  
QoS policies consist of a collection of user defined traffic classes. A policy  
can be assigned to more than one port, but a port may only have one  
policy.  
Note that the switch can only perform error checking of parameters and  
parameter values for the policy and its traffic classes and flow groups  
when the policy is set on a port.  
QoS controls are applied to ingress traffic on ports. Therefore, to control a  
particular type of traffic, an appropriate QoS policy must be attached to  
each port that type of traffic ingresses.  
Although a policy can be applied to an egress port, the classifiers and the  
QoS controls are actually applied by the switch on the ingress ports of the  
traffic. This means the parameters used to classify the traffic and the  
actions specified by the policy are checked and applied on the ingress  
traffic of every port, before the traffic reaches an egress queue. As a  
consequence, a policy is never applied to the whole aggregated traffic of a  
designated egress port, but rather to the individual ingress flows destined  
to the port.  
The effects of this behavior become evident when using the maximum  
bandwidth feature of QoS. Here is an example. Suppose you have a policy  
that assigns 5 Mbps of maximum bandwidth to an egress port. Now  
assume there are 10 ports on the switch where ingress traffic matches the  
criteria specified in the classifier assigned to the policy of the egress port.  
Since the policy considers each ingress flow separately, the result would  
be a maximum bandwidth of 50 Mbps (10 x 5 Mbps) on the egress port,  
because there are 10 flows, one from each ingress port, directed to the  
egress port.  
An additional factor to consider when specifying an egress port in a policy  
is that if the destination MAC address of the traffic flow has not been  
learned by the egress port or, alternatively, added as a static address to  
the port, the policy remains inactive. This is because the ingress ports  
consider the traffic as unknown traffic and flood the traffic to all the ports.  
This applies equally to unknown unicast and unknown multicast traffic, as  
well as broadcast traffic.  
146  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
QoS Policy Guidelines  
Following is a list of QoS policy guidelines:  
ˆ A classifier may be assigned to many flow groups. However, assigning  
a classifier more than once within the same policy may lead to  
undesirable results. A classifier may be used successfully in many  
different policies.  
ˆ A flow group must be assigned at least one classifier but may have  
many classifiers.  
ˆ A flow group may be assigned to only one traffic class.  
ˆ A traffic class may have many flow groups.  
ˆ A traffic class may only be assigned to one policy.  
ˆ A policy may have many traffic classes.  
ˆ A policy may be assigned to many ports.  
ˆ A port may only have one policy.  
ˆ You can create a policy without assigning it to a port, but the policy will  
be inactive.  
ˆ A policy must have at least one action defined in the flow group, traffic  
class, or the policy itself. A policy without an action is invalid.  
ˆ A Quality of Service policy and an access control list can coexist on the  
same port only if they have different classifiers.  
ˆ The switch can store up to 64 flow groups.  
ˆ The switch can store up to 64 traffic classes.  
ˆ The switch can store up to 64 policies.  
Section II: Advanced Operations  
147  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 13: Quality of Service  
Packet Processing  
You can use the switch’s QoS tools to perform any combination of the  
following functions on a packet flow:  
ˆ Limiting bandwidth  
ˆ Prioritizing packets to determine the level of precedence the switch will  
give to the packet for processing  
ˆ Replacing the VLAN tag User Priority to enable the next switch in the  
network to process the packet correctly  
ˆ Replacing the TOS precedence or DSCP value to enable the next  
switch in the network to process the packet correctly.  
Bandwidth Allocation  
Bandwidth limiting is configured at the level of traffic classes, and  
encompasses the flow groups contained in the traffic class. Traffic classes  
can be assigned maximum bandwidths, specified in kbps, Mbps, or Gbps.  
Packet Prioritization  
The switch has eight Class of Service (CoS) egress queues, numbered  
from 0 to 7. Queue 7 has the highest priority. When the switch becomes  
congested, it gives high priority queues precedence over lower-priority  
queues. When the switch has information about a packet’s priority, it  
sends the packet to the appropriate queue. You can specify the queue  
where the switch sends traffic, how much precedence each queue has,  
and whether priority remapping is written into the packet’s header for the  
next hop to use.  
Prioritizing packets cannot improve your network’s performance when  
bandwidth is over-subscribed to the point that egress queues are always  
full. If one type of traffic is causing the congestion, you can limit its  
bandwidth. Other solutions are to increase bandwidth or decrease traffic.  
You can set a packet’s priority by configuring a priority in the flow group or  
traffic class to which the packet belongs. The packet is put in the  
appropriate CoS queue for that priority. If the flow group and traffic class  
do not include a priority, the switch can determine the priority from the  
VLAN tag User Priority field of incoming tagged packets. The packet is put  
in the appropriate CoS queue for its VLAN tag User Priority field. If neither  
the traffic class / flow group priority nor the VLAN tag User Priority is set,  
the packet is sent to the default queue, queue 1.  
148  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Both the VLAN tag User Priority and the traffic class / flow group priority  
setting allow eight different priority values (0-7). These eight priorities are  
mapped to the switch’s eight CoS queues. The switch’s default mapping is  
shown in Table 10 on page 134. Note that priority 0 is mapped to CoS  
queue 1 instead of CoS queue 0 because tagged traffic that has never  
been prioritized has a VLAN tag User Priority of 0. If priority 0 was mapped  
to CoS queue 0, this default traffic goes to the lowest queue, which is  
probably undesirable. This mapping also makes it possible to give some  
traffic a lower priority than the default traffic.  
Section II: Advanced Operations  
149  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 13: Quality of Service  
Replacing Priorities  
The traffic class or flow group priority (if set) determines the egress queue  
a packet is sent to when it egresses the switch, but by default has no  
effect on how the rest of the network processes the packet. To  
permanently change the packet’s priority, you need to replace one of two  
priority fields in the packet header:  
ˆ The User Priority field of the VLAN tag header. Replacing this field  
relabels VLAN-tagged traffic, so that downstream switches can  
process it appropriately.  
ˆ The DSCP value of the IP header’s TOS byte (Figure 5 on page 115).  
Replacing this field may be required as part of the configuration of a  
DiffServ domain. See “DiffServ Domains” on page 151 for information  
on using the QoS policy model and the DSCP value to configure a  
DiffServ domain.  
VLAN Tag User Priorities  
Within a flow group or traffic class, the VLAN tag User Priority value of  
incoming packets can be replaced with the priority specified in the flow  
group or traffic class. Replacement occurs before the packet is queued, so  
this priority also sets the queue priority.  
DSCP Values  
There are three methods for replacing the DSCP byte of an incoming  
packet. You can use these methods together or separately. They are  
described in the order in which the switch performs them.  
ˆ The DSCP value can be overwritten at ingress, for all traffic in a policy.  
ˆ The DSCP value in the packet can be replaced at the traffic class or  
flow group level.  
ˆ You can use these two replacements together at the edge of a DiffServ  
domain, to initialize incoming traffic.  
ˆ The DSCP value in a flow of packets can replaced if the bandwidth  
allocated to that traffic class is exceeded. This option allows the next  
switch in the network to identify traffic that exceeded the bandwidth  
allocation.  
150  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
DiffServ Domains  
Differentiated Services (DiffServ) is a method of dividing IP traffic into  
classes of service, without requiring that every router in a network  
remember detailed information about traffic flows. DiffServ operates within  
a DiffServ domain, a network or subnet that is managed as a single QoS  
unit. Packets are classified according to user-specified criteria at the edge  
of the network, divided into classes, and assigned the required class of  
service. Then packets are marked with a Differentiated Services Code  
Point (DSCP) tag to indicate the class of service to which they belong. The  
DSCP value is written into the TOS field of the IP header. Routers within  
the network then use this DSCP value to classify packets and assign QoS  
appropriately. When a packet leaves the DiffServ domain, the DSCP value  
can be replaced with a value appropriate for the next DiffServ domain.  
A simple example of this process is shown in Figure 12, for limiting the  
amount of bandwidth used by traffic from a particular IP address. In the  
domain shown, this bandwidth limit is supplied by the class of service  
represented by a DSCP value of 40. In the next DiffServ domain, this  
traffic is assigned to the class of service represented by a DSCP value of  
3.  
DiffServ Domain  
Classify by source IP address  
Mark with DSCP=40  
Limit bandwidth  
Classify by DSCP=40  
Limit bandwidth  
Re-mark to DSCP=3  
Non-DiffServ  
traffic  
Next DiffServ  
domain  
Classify by DSCP=40  
Limit bandwidth  
Figure 12. DiffServ Domain Example  
Section II: Advanced Operations  
151  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 13: Quality of Service  
To use the QoS tool set to configure a DiffServ domain:  
1. As packets come into the domain at edge switches, replace their  
DSCP value, if required.  
ˆ Classify the packets according to the required characteristics. For  
ˆ Assign the classifiers to flow groups and the flow groups to traffic  
classes, with a different traffic class for each DiffServ code point  
grouping within the DiffServ domain.  
ˆ Give each traffic class the priority and/or bandwidth limiting  
controls that are required for that type of packet within this part of  
the domain.  
ˆ Assign a DSCP value to each traffic class, to be written into the  
TOS field of the packet header.  
2. On switches and routers within the DiffServ domain, classify packets  
according to the DSCP values that were assigned to traffic classes on  
the edge switches.  
ˆ Assign the classifiers to flow groups and the flow groups to traffic  
classes, with a different traffic class for each DiffServ code point  
grouping within the DiffServ domain.  
ˆ Give each traffic class the priority and/or bandwidth limiting  
controls that are required for that type of packet within this part of  
the domain. These QoS controls need not be the same for each  
switch.  
3. As packets leave the DiffServ domain, classify them according to the  
DSCP values.  
ˆ Assign the classifiers to flow groups and the flow groups to traffic  
classes, with a different traffic class for each DiffServ code point  
grouping within the DiffServ domain.  
ˆ Give each traffic class the priority and/or bandwidth limiting  
controls required for transmission of that type of packet to its next  
destination, in accordance with any Service Level Agreement  
(SLA) with the providers of that destination.  
ˆ If necessary, assign a different DSCP value to each traffic class, to  
be written into the TOS field of the packet header, to match the  
DSCP or TOS priority values of the destination network.  
152  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Examples  
The following examples demonstrate how to implement QoS in three  
situations:  
Voice Voice applications typically require a small but consistent bandwidth. They  
are sensitive to latency (interpacket delay) and jitter (delivery delay). Voice  
Applications  
applications can be set up to have the highest priority.  
This example creates two policies that ensure low latency for all traffic sent  
by and destined to a voice application located on a node with the IP  
address 149.44.44.44. The policies raise the priority level of the packets to  
7, the highest level. Policy 6 is for traffic from the application that enter the  
switch on port 1. Policy 11 is for traffic arriving on port 8 going to the  
application. The components of the policies are shown in Figure 13.  
Section II: Advanced Operations  
153  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 13: Quality of Service  
Policy 6  
Policy 11  
Create Classifier  
Create Classifier  
01 - Classifier ID: ..... 23  
02 - Description ....... VoIP flow  
.
01 - Classifier ID: ..... 22  
02 - Description ....... VoIP flow  
.
14 - Dst IP Addr ....... 149.44.44.44  
15 - Dst IP Mask .......  
12 - Src IP Addr ....... 149.44.44.44  
13 - Src IP Mask ......  
Create Flow Group  
Create Flow Group  
1 - Flow Group ID ............. 17  
2 - Description ................... VoIP  
3 - DSCP Value .................  
4 - Priority ......................... 7  
5 - Remark Priority ........... No  
.
1 - Flow Group ID ............. 14  
2 - Description ................... VoIP  
3 - DSCP Value .................  
4 - Priority ......................... 7  
5 - Remark Priority ........... No  
.
9 - Classifier List .............. 23  
9 - Classifier List .............. 22  
Create Traffic Class  
Create Traffic Class  
01 - Traffic Class ID: ........ 18  
02 - Desciption ................ VoIP flow  
.
01 - Traffic Class ID: ........ 15  
02 - Desciption ................ VoIP flow  
.
E - Flow Group List ......... 14  
E - Flow Group List ......... 17  
Create Policy  
Create Policy  
1 - Policy ID: .................. 6  
1 - Policy ID: .............. 11  
2 - Desciption ................ VoIP flow  
2 - Desciption ............ VoIP flow  
.
.
9 - Traffic Class List ......... 18  
9 - Traffic Class List ..... 15  
.
.
B - Ingress Port List ......... 1  
B - Ingress Port List ...... 8  
Figure 13. QoS Voice Application Example  
The parts of the policies are:  
ˆ Classifier - Defines the traffic flow by specifying the IP address of the  
node with the voice application. The classifier for Policy 6 specifies the  
address as a source address because this classifier is part of a policy  
for packets coming from the application. The classifier for Policy 11  
specifies the address as a destination address because this classifier  
is part of a policy for packets going to the application.  
ˆ Flow Group - Specifies the new priority level of 7 for the packets. In this  
example the packets leave the switch with the same priority level they  
had when they entered. The new priority level is relevant only as the  
packets traverse the switch. To change the packets’ priority level so  
that they leave with the new level, you would change option 5, Remark  
Priority, to Yes.  
154  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
ˆ Traffic Class - No action is taken by the traffic class, other than to  
specify the flow group. Traffic class has a priority setting you can use  
to override the priority level of packets, just as in a flow group. If you  
enter a priority value in both places, the setting in the flow group  
overrides the setting in the traffic class.  
ˆ Policy - Specifies the traffic class and the port to which the policy is to  
be assigned. Policy 6 is applied to port 1 because this is where the  
application is located. Policy 11 is applied to port 8 because this is  
where traffic going to the application will be received.  
Video Video applications typically require a larger bandwidth than voice  
applications. Video applications can be set up to have a high priority and  
buffering, depending on the application.  
Applications  
This example creates policies with low latency and jitter for video streams  
(for example, net conference calls). The policies in Figure 14 assign the  
packets a priority level of 4. The policies also limit the bandwidth for the  
video streams to 5 Mbps to illustrate how you can combine a change to the  
priority level with bandwidth restriction to further define traffic control. The  
node containing the application has the IP address 149.44.44.44. Policy  
17 is assigned to port 1, where the application is located, and Policy 32 is  
assigned to port 8 where packets destined to the application enter the  
switch.  
Section II: Advanced Operations  
155  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 13: Quality of Service  
Policy 17  
Policy 32  
Create Classifier  
Create Classifier  
01 - Classifier ID: ..... 16  
02 - Desciption ......... Video flow  
.
12 - Src IP Addr ....... 149.44.44.44  
13 - Src IP Mask .......  
01 - Classifier ID: ..... 42  
02 - Desciption ......... Video flow  
.
12 - Dst IP Addr ........ 149.44.44.44  
13 - Dst IP Mask .......  
Create Flow Group  
Create Flow Group  
1 - Flow Group ID ............. 41  
2 - Description ................... Video  
3 - DSCP Value .................  
4 - Priority ......................... 4  
5 - Remark Priority ........... No  
.
1 - Flow Group ID ............. 36  
2 - Description ................... Video  
3 - DSCP Value .................  
4 - Priority ......................... 4  
5 - Remark Priority ........... No  
.
9 - Classifier List .............. 16  
9 - Classifier List .............. 42  
Create Traffic Class  
Create Traffic Class  
1 - Traffic Class ID: ........ 19  
2 - Desciption ................ Video  
.
6 - Max Bandwidth ........ 5  
.
E - Flow Group List ....... 41  
1 - Traffic Class ID: ........ 21  
2 - Desciption ................ Video  
.
6 - Max Bandwidth ........ 5  
.
E - Flow Group List ....... 36  
Create Policy  
Create Policy  
1 - Policy ID: ................ 17  
1 - Policy ID: ................ 32  
2 - Desciption .............. Video flow  
2 - Desciption .............. Video flow  
.
.
9 - Traffic Class List ....... 19  
9 - Traffic Class List ....... 21  
.
.
B - Ingress Port List ....... 1  
B - Ingress Port List ....... 8  
Figure 14. QoS Video Application Example  
The parts of the policies are:  
ˆ Classifier - Specifies the IP address of the node with a video  
application. The classifier for Policy 17 specifies the address as a  
source address since this classifier is part of a policy concerning  
packets coming from the application. The classifier for Policy 32  
specifies the address as a destination address because this classifier  
is part of a policy concerning packets going to the application.  
ˆ Flow Group - Specifies the new priority level of 4 for the packets. As  
with the previous example, the packets leave the switch with the same  
priority level they had when they entered. The new priority level is  
relevant only while the packets traverse the switch. To alter the  
156  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
packets so they leave containing the new level, you would change  
option 5, Remark Priority, to Yes.  
ˆ Traffic Class - The packet stream is assigned a maximum bandwidth of  
5 Mbps. Bandwidth assignment can only be made at the traffic class  
level.  
ˆ Policy - Specifies the traffic class and the port where the policy is to be  
assigned.  
Critical Database Critical databases typically require a high bandwidth. They also typically  
require less priority than either voice or video.  
The policies in Figure 15 assign 50 Mbps bandwidth, with no change to  
priority, to traffic going to and from a database. The database is located on  
a node with the IP address 149.44.44.44 on port 1 of the switch.  
Policy 15  
Policy 17  
Create Classifier  
Create Classifier  
01 - Classifier ID: ..... 42  
02 - Description ....... Database  
.
12 - Src IP Addr ...... 149.44.44.44  
13 - Src IP Mask .....  
01 - Classifier ID: ..... 10  
02 - Description ........ Database  
.
14 - Dst IP Addr ....... 149.44.44.44  
15 - Dst IP Mask ......  
Create Flow Group  
Create Flow Group  
1 - Flow Group ID ............. 36  
2 - Description ................... Database  
3 - DSCP Value .................  
4 - Priority .........................  
5 - Remark Priority ........... No  
.
1 - Flow Group ID ............. 12  
2 - Description ................... Database  
3 - DSCP Value .................  
4 - Priority .........................  
5 - Remark Priority ........... No  
.
9 - Classifier List .............. 42  
9 - Classifier List .............. 10  
Create Traffic Class  
Create Traffic Class  
1 - Traffic Class ID: ........ 21  
2 - Description ............... Database  
.
6 - Max Bandwidth ........ 50  
.
E - Flow Group List ....... 36  
1 - Traffic Class ID: ........ 17  
2 - Description ............... Database  
.
6 - Max Bandwidth ........ 50  
.
E - Flow Group List ....... 12  
Create Policy  
Create Policy  
1 - Policy ID: ................ 15  
2 - Description ............. Database  
.
9 - Traffic Class List ....... 21  
.
B - Ingress Port List ....... 1  
1 - Policy ID: ................ 17  
2 - Description ............. Database  
.
9 - Traffic Class List ....... 17  
.
B - Ingress Port List ....... 8  
Figure 15. QoS Critical Database Example  
Section II: Advanced Operations  
157  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 13: Quality of Service  
Policy The purpose of this example is to illustrate the hierarchy of the  
components of a QoS policy and how that hierarchy needs to be taken into  
account when assigning new priority and DSCP values. A new priority can  
be set at the flow group and traffic class levels, while a new DSCP value  
can be set at all three levels—flow group, traffic class and policy. The  
basic rules are:  
Component  
Hierarchy  
ˆ A new setting in a flow group takes precedence over a corresponding  
setting in a traffic class or policy.  
ˆ A new setting in a traffic class takes precedence over a corresponding  
setting in a policy.  
ˆ A new setting in a policy is used only if there is no corresponding  
setting in a flow group or traffic class.  
This concept is illustrated in Figure 16 on page 159. It shows a policy for a  
series of traffic flows consisting of subnets defined by their destination IP  
addresses. New DSCP values for the traffic flows are established at  
different levels within the policy.  
Traffic flows 149.11.11.0 and 149.22.22.0, defined by classifiers 1 and 2,  
are attached to a flow group, traffic class, and policy that contain new  
DSCP values. Because a setting in a flow group takes precedence over  
that of a traffic class or policy, the value in the flow group is used. The  
result is that the DSCP value in the two traffic flows is changed to 10.  
The flow group for traffic flows 149.33.33.0 and 149.44.44.0, defined in  
classifiers 3 and 4, does not contain a new DSCP value. Therefore, the  
new value in the traffic class is used, in this case 30. The policy also has a  
DSCP setting, but it is not used for these traffic flows because a new  
DSCP setting in a traffic class takes precedence over that of a policy.  
Finally, the new DSCP value for traffic flows 149.55.55.0 and 149.66.66.0,  
defined in classifiers 5 and 6, is set at the policy level to a value of 55  
because the flow group and traffic class do not specify a new value.  
158  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Create Classifier  
01 - Classifier ID: ..... 1  
.
14 - Dst IP Addr ..... 149.11.11.0  
15 - Dst IP Mask ..... 255.255.255.0  
Create Flow Group  
1 - Flow Group ID ......... 1  
.
3 - DSCP Value ............. 10  
.
Create Classifier  
9 - Classifier List ............1,2  
01 - Classifier ID: ..... 2  
.
Create Traffic Class  
14 - Dst IP Addr ..... 149.22.22.0  
15 - Dst IP Addr ...... 255.255.255.0  
1 - Traffic Class ID: ........ 1  
.
5 - DSCP value ............. 30  
.
Create Classifier  
E - Flow Group List ....... 1,2  
01 - Classifier ID: ..... 3  
.
14 - Dst IP Addr ..... 149.33.33.0  
15 - Dst IP Mask .... 255.255.255.0  
Create Flow Group  
1 - Flow Group ID ......... 2  
.
3 - DSCP Value .............  
.
Create Policy  
Create Classifier  
9 - Classifier List ............3,4  
1 - Policy ID: ................ 1  
.
3 - Remark DSCP ........ All  
4 - DSCP value ............ 55  
.
01 - Classifier ID: ..... 4  
.
14 - Dst IP Addr ....... 149.44.44.0  
15 - Dst IP Addr ....... 255.255.255.0  
9 - Traffic Class List ..... 1,2  
Create Classifier  
01 - Classifier ID: ..... 5  
.
14 - Dst IP Addr ....... 149.55.55.0  
15 - Dst IP Mask ...... 255.255.255.0  
Create Flow Group  
Create Traffic Class  
1 - Flow Group ID ......... 3  
1 - Traffic Class ID: ........ 2  
.
.
3 - DSCP Value .............  
5 - DSCP value .............  
Create Classifier  
.
.
01 - Classifier ID: ..... 6  
.
9 - Classifier List ............5,6  
E - Flow Group List ....... 3  
14 - Dst IP Addr ..... 149.66.66.0  
15 - Dst IP Mask ...... 255.255.255.0  
Figure 16. Policy Component Hierarchy Example  
Section II: Advanced Operations  
159  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 13: Quality of Service  
160  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 14  
Denial of Service Defenses  
This chapter explains the defense mechanisms in the management  
software that can protect your network against denial of service (DoS)  
attacks. Sections in the chapter include:  
Section II: Advanced Operations  
161  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 14: Denial of Service Defenses  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
162  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
The AT-S63 Management Software can help protect your network against  
the following types of denial of service attacks.  
ˆ SYN Flood Attack  
ˆ Smurf Attack  
ˆ Land Attack  
ˆ Teardrop Attack  
ˆ Ping of Death Attack  
ˆ IP Options Attack  
The following sections describe each type of attack and the mechanism  
employed by the AT-S63 Management Software to protect your network.  
Note  
Be sure to read the following descriptions before implementing a  
DoS defense on a switch. Some defense mechanisms are CPU  
intensive and can impact switch behavior.  
Section II: Advanced Operations  
163  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 14: Denial of Service Defenses  
SYN Flood Attack  
In this type of attack, an attacker sends a large number of TCP connection  
requests (TCP SYN packets) with bogus source addresses to the victim.  
The victim responds with acknowledgements (SYN ACK packets), but  
because the original source addresses are bogus, the victim node does  
not receive any replies. If the attacker sends enough requests in a short  
enough period, the victim may freeze operations when the number of  
requests exceeds the capacity of its connections queue.  
To defend against this form of attack, a switch port monitors the number of  
ingress TCP connection requests it receives. If a port receives more than  
60 requests per second, the following occurs.  
ˆ The switch sends an SNMP trap to the management stations  
ˆ The switch port is blocked for one minute.  
This defense mechanism does not involve the switch’s CPU. You can  
activate it on some or all of the ports without impacting switch  
performance.  
164  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Smurf Attack  
This DoS attack is instigated by an attacker sending a ICMP Echo (Ping)  
request that has the network’s IP broadcast address as the destination  
address and the address of the victim as the source of the ICMP Echo  
(Ping) request. This overwhelms the victim with a large number of ICMP  
Echo (Ping) replies from the other network nodes.  
A switch port defends against this form of attack by examining the  
destination IP addresses of ingress ICMP Echo (Ping) request packets  
and discarding those that contain the network’s IP broadcast address as a  
destination address.  
To implement this defense, you must specify an IP address of a node on  
your network and a mask. The switch uses the two to determine the  
broadcast address of your network.  
This defense mechanism does not involve the switch’s CPU. You can  
activate it on some or all of the ports without impacting switch  
performance.  
Section II: Advanced Operations  
165  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 14: Denial of Service Defenses  
Land Attack  
In this attack, an attacker sends a bogus IP packet where the source and  
destination IP addresses are the same. This leaves the victim thinking that  
it is sending a message to itself.  
The most direct approach for defending against this form of attack is for  
the AT-S63 Management Software to check the source and destination IP  
addresses in the IP packets, searching for and discarding those with  
identical source and destination addresses. However, this would require  
too much processing by the switch’s CPU and would adversely impact  
switch performance.  
Instead, the switch examines the IP packets that are entering and leaving  
your network. IP packets that are generated within your network and  
contain a local IP address as the destination address are not allowed to  
leave the network, and IP packets that are generated outside the network  
but contain a local IP address as the source address are not allowed into  
the network.  
In order for this defense mechanism to work, you need to specify an uplink  
port. This is the port on the switch that is connected to a device, such as a  
DSL router, that leads outside your network. You can specify only one  
uplink port.  
Note  
You should not use this defense mechanism on a switch that is not  
connected to a device that leads outside your network.  
You also need to enter the IP address of one of your network devices as  
well as a mask which the switch uses to differentiate between the network  
portion and node portion of the address. The switch uses the IP address  
and mask to determine which IP addresses are local to your network and  
which are from outside you network.  
The following is a overview of how the process works. This example  
assumes that you have activated the feature on port 4, which is connected  
to a device local to your network, and that you specified port 1 as the  
uplink port, which is connected to the device that leads outside your  
network. The steps below review what happens when an ingress IP  
packet from the local device arrives on port 4:  
1. When port 4 receives an ingress IP packet with a destination MAC  
address learned on uplink port 1, it examines the packet’s source IP  
address.  
166  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
2. If the source IP address is not local to the network, it discards the  
packet because it assumes that a packet with an IP address that is not  
local to the network should not be appearing on a port that is not an  
uplink port. This protects against the possibility of a Land attack  
originating from within your network.  
3. If the source IP address is local to the network, the port forwards the  
packet to uplink port 1.  
Below is a review of how the process takes place when an ingress IP  
packet arrives on uplink port 1 that is destined for port 4:  
1. When uplink port 1 receives an ingress IP packet with a destination  
MAC address that was learned on port 4, it examines the packet’s  
source IP address before forwarding the packet.  
2. If the source IP address is local to the network, uplink port 1 does not  
forward the packet to port 4 because it assumes that a packet with a  
source IP address that is local to the network should not be entering  
the network from outside the network on the uplink port.  
3. If the source IP address is not local to the network, port 1 forwards the  
packet to port 4.  
The following guidelines apply to using this defense mechanism:  
ˆ If you choose to use it, Allied Telesis recommends activating it on all  
ports on the switch, including the uplink port.  
ˆ You can specify only one uplink port.  
ˆ You must specify the IP address of one of the network nodes,  
preferably the lowest IP address, and a mask.  
This form of defense is not CPU intensive. Activating it on all ports should  
not affect switch behavior.  
Section II: Advanced Operations  
167  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 14: Denial of Service Defenses  
Teardrop Attack  
An attacker sends an IP packet in several fragments with a bogus offset  
value, used to reconstruct the packet, in one of the fragments to a victim.  
Because of the bogus offset value, the victim is unable to reassemble the  
packet, possibly causing it to freeze operations.  
The defense mechanism for this type of attack has all ingress fragmented  
IP traffic received on a port sent to the switch’s CPU. The CPU samples  
related, consecutive fragments, checking for fragments with invalid offset  
values.  
If one is found, the following occurs:  
ˆ The switch sends an SNMP trap to the management stations.  
ˆ The switch port is blocked for one minute.  
Because the CPU only samples the ingress IP traffic, this defense  
mechanism may not catch all occurrences of this form of attack.  
Caution  
This defense is extremely CPU intensive; use with caution.  
Unrestricted use can cause a switch to halt operations if the CPU  
becomes overwhelmed with IP traffic. To prevent this, Allied Telesis  
recommends activating this defense on only the uplink port and one  
other switch port at a time.  
168  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Ping of Death Attack  
The attacker sends an oversized, fragmented ICMP Echo (Ping) request  
(greater than 65,535 bits) to the victim, which, if lacking a policy for  
handling oversized packets, may freeze.  
To defend against this form of attack, a switch port searches for the last  
fragment of a fragmented ICMP Echo (Ping) request and examines its  
offset to determine if the packet size is greater than 63,488 bits. If it is, the  
fragment is forwarded to the switch’s CPU for final packet size  
determination. If the switch determines that the packet is oversized, the  
following occurs:  
ˆ The switch sends an SNMP trap to the management stations.  
ˆ The switch port is blocked for one minute.  
Note  
This defense mechanism requires some involvement by the switch’s  
CPU, though not as much as the Teardrop defense. This does not  
impact the forwarding of traffic between the switch ports, but it can  
affect the handling of CPU events, such as the processing of IGMP  
packets and spanning tree BPDUs. For this reason, Allied Telesis  
recommends limiting the use of this defense, activating it only on  
those ports where an attack is most likely to originate.  
Also note that an attacker can circumvent the defense by sending a  
stream of ICMP Echo (Ping) requests with a size of 63,488 to 65,534 bits.  
A large number of requests could overwhelm the switch’s CPU.  
Section II: Advanced Operations  
169  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 14: Denial of Service Defenses  
IP Options Attack  
In the basic scenario of an IP attack, an attacker sends packets containing  
bad IP options. There are several types of IP option attacks and the  
AT-S63 Management Software does not distinguish between them.  
Rather, the defense mechanism counts the number of ingress IP packets  
containing IP options received on a port. If the number exceeds 20  
packets per second, the switch considers this a possible IP options attack  
and the following occurs:  
ˆ It sends an SNMP trap to the management stations.  
ˆ The switch port is blocked for one minute.  
This defense mechanism does not involve the switch’s CPU. You can  
activate it on as many ports as you want without it impacting switch  
performance.  
Note  
This defense does not actually check IP packets for bad IP options,  
and so can only alert you to a possible attack.  
170  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Mirroring Traffic  
The Land, Teardrop, Ping of Death, and IP Options defense mechanisms  
allow you to copy the examined traffic to a mirror port for further analysis  
with a data sniffer or analyzer. This feature differs slightly from port  
mirroring in that prior to an actual violation of a defense mechanism, only  
the packets examined by a defense mechanism, rather than all packets,  
are mirrored to the destination port. Should a violation occur, then all  
ingress packets on the port where the violation occurred are mirrored.  
As an example, activating the mirroring feature in conjunction with the  
Teardrop defense on a port sends all examined ingress fragmented IP  
traffic to the destination mirror port. If the switch detects a violation, all  
ingress packets on the port are copied to the mirror port during the sixty  
seconds that the port is blocked.  
Implementing this feature requires configuring the port mirroring feature as  
follows:  
ˆ Activate port mirroring.  
ˆ Specify a destination port.  
ˆ Do not specify any source ports. The source ports are defined by the  
Denial of Service defense mechanism.  
Section II: Advanced Operations  
171  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 14: Denial of Service Defenses  
Denial of Service Defense Guidelines  
Below are guidelines to observe when using this feature:  
ˆ A switch port can support more than one DoS defense at a time.  
ˆ The Teardrop and the Ping of Death defenses are CPU intensive. Use  
these defenses with caution.  
172  
Section II: Advanced Operations  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Section III  
Snooping Protocols  
The chapters in this section contain overview information on the snooping  
protocols. The chapters include:  
Section III: Snooping Protocols  
173  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
174  
Section III: Snooping Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 15  
IGMP Snooping  
This chapter explains Internet Group Management Protocol (IGMP)  
snooping feature in the following sections:  
Section III: Snooping Protocols  
175  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 15: IGMP Snooping  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
176  
Section III: Snooping Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
IPv4 routers use IGMP to create lists of nodes that are members of  
multicast groups. (A multicast group is a group of end nodes that want to  
receive multicast packets from a multicast application.) The router creates  
a multicast membership list by periodically sending out queries to the local  
area networks connected to its ports.  
A node wanting to become a member of a multicast group responds to a  
query by sending a report. A report indicates an end node’s desire to  
become a member of a multicast group. Nodes that join a multicast group  
are referred to as host nodes. After becoming a member of a multicast  
group, a host node must continue to periodically issue reports to remain a  
member.  
After the router has received a report from a host node, it notes the  
multicast group that the host node wants to join and the port on the router  
where the node is located. Any multicast packets belonging to that  
multicast group are then forwarded by the router out the port. If a particular  
port on the router has no nodes that want to be members of multicast  
groups, the router does not send multicast packets out the port. This  
improves network performance by restricting multicast packets only to  
router ports where host nodes are located.  
There are three versions of IGMP — versions 1, 2, and 3. One of the  
differences between the versions is how a host node signals that it no  
longer wants to be a member of a multicast group. In version 1 it stops  
sending reports. If a router does not receive a report from a host node after  
a predefined length of time, referred to as a time-out value, it assumes that  
the host node no longer wants to receive multicast frames, and removes it  
from the membership list of the multicast group.  
In version 2 a host node exits from a multicast group by sending a leave  
request. After receiving a leave request from a host node, the router  
removes the node from appropriate membership list. The router also stops  
sending multicast packets out the port to which the node is connected if it  
determines there are no further host nodes on the port.  
Version 3 adds the ability of host nodes to join or leave specific sources in  
a multicast group.  
The IGMP snooping feature on the AT-9400 Switch supports all three  
versions of IGMP. The switch monitors the flow of queries from routers  
and reports and leave messages from host nodes to build its own multicast  
membership lists. It uses the lists to forward multicast packets only to  
switch ports where there are host nodes that are members of multicast  
groups. This improves switch performance and network security by  
restricting the flow of multicast packets only to those switch ports  
connected to host nodes.  
Section III: Snooping Protocols  
177  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 15: IGMP Snooping  
Without IGMP snooping a switch would have to flood multicast packets out  
all of its ports, except the port on which it received the packet. Such  
flooding of packets can negatively impact network performance.  
The AT-9400 Switch maintains its list of multicast groups through an  
adjustable timeout value, which controls how frequently it expects to see  
reports from end nodes that want to remain members of multicast groups,  
and by processing leave requests.  
Note  
The default setting for IGMP snooping on the switch is disabled.  
178  
Section III: Snooping Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 16  
MLD Snooping  
This chapter explains Multicast Listener Discovery (MLD) snooping:  
Section III: Snooping Protocols  
179  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 16: MLD Snooping  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from the following management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
180  
Section III: Snooping Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
MLD snooping performs the same function as IGMP snooping. The switch  
uses the feature to build multicast membership lists. It uses the lists to  
forward multicast packets only to switch ports where there are host nodes  
that are members of the multicast groups. The difference between the two  
is that MLD snooping is for IPv6 and IGMP snooping for IPv4  
environments. (For background information on IGMP snooping, refer to  
There are two versions of MLD. MLDv1 is equivalent to IGMPv2 and  
MLDv2 is equivalent to IGMPv3. The AT-9400 Switch supports snooping  
of both MLDv1 and MLDv2.  
Note  
The default setting for MLD snooping on the switch is disabled.  
Section III: Snooping Protocols  
181  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 16: MLD Snooping  
182  
Section III: Snooping Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 17  
RRP Snooping  
This chapter explains RRP snooping and contains the following sections:  
Section III: Snooping Protocols  
183  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 17: RRP Snooping  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from the following management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
184  
Section III: Snooping Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
The Router Redundancy Protocol (RRP) allows multiple routers to share  
the same virtual IP address and MAC address. In network topologies  
where redundant router paths or links exist, the protocol enables routers,  
through an election process, to designate one as the master router. This  
router functions as the provider of the primary path between LAN  
segments. Slave routers function as backup paths in the event that the  
master router or primary path fails.  
Because the master and slave routers are able to share the same virtual  
IP address and MAC address, a change in data paths need not  
necessitate an adjustment to the default gateways on the network nodes  
that employ the routers. When a slave router transitions to master, it uses  
the same IP address as the previous master router, making the transition  
transparent to the network end nodes. In large networks, these  
transparent transitions can save the time and effort of having to manually  
reconfigure default gateway addresses on large numbers of network  
nodes when a router pathway fails.  
RRP snooping on the AT-9400 Switch facilitates the transition to a new  
master router by minimizing the loss of traffic, and so reduces the impact  
the transition could have on your network traffic. RRP snooping monitors  
ingress RRP packets, determined by their source MAC address. Source  
MAC addresses considered by the AT-S63 Management Software as RRP  
packets are:  
ˆ 00:E0:2B:00:00:80-9F  
ˆ 00:A0:D2EB:FF:00  
ˆ 00:00:5E:00:01:00-FF  
A port receiving an RRP packet is deemed by the switch as the master  
RRP port. The virtual MAC address of the router is entered as a dynamic  
address on the port. If the switch starts to receive RRP packets on another  
port, it assumes that a backup or slave router has made the transition to  
the role of the new master router.  
The switch responds by deleting all dynamic MAC addresses from the  
MAC address table. As the switch relearns the addresses, the virtual MAC  
address of the new master router is learned on the new master RRP port,  
rather than the old port. Any packets received by the switch and destined  
for the router are forwarded to the new master router.  
Section III: Snooping Protocols  
185  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 17: RRP Snooping  
Guidelines  
The following guidelines apply to the RRP snooping feature:  
ˆ The default setting for this feature is disabled.  
ˆ Activating the feature flushes all dynamic MAC addresses from the  
MAC address table.  
ˆ RRP snooping is supported on ports operating in the MAC address-  
based port security level of automatic. This feature is not supported on  
ports operating with a security level of limited, secured, or locked.  
ˆ RRP snooping is supported on port trunks.  
186  
Section III: Snooping Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 18  
Ethernet Protection Switching Ring  
Snooping  
This chapter has the following sections:  
Section III: Snooping Protocols  
187  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 18: Ethernet Protection Switching Ring Snooping  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
Not supported.  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature must be managed from the command line interface.  
188  
Section III: Snooping Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
Ethernet Protection Switching Ring is a feature found on selected Allied  
Telesis products, such as the AT-8948 Series Gigabit Layer 3 Switches. It  
offers an effective alternative to spanning tree based options when using  
ring based topologies to create high speed resilient networks.  
EPSR consists of a master node and a number of transit nodes in a ring  
configuration. The master node monitors the health of the ring by  
transmitting healthcheck messages from a primary port at regular intervals  
over a control VLAN, and watching for the messages on a secondary port.  
If the healthcheck messages fail to arrive, the master node commences  
fault recovery of the ring by activating the secondary port so that  
connectivity between the transit nodes is maintained through the master  
node. When the integrity of the ring is restored, and the healthcheck  
messages can again traverse the entire ring, the master switch returns the  
secondary port to the blocking state.  
Note  
For background information and configuration examples of EPSR,  
refer to the AlliedWare OS Software Reference Guide.  
EPSR snooping gives the AT-9400 Switch the ability to function as a  
transit node of a ring, but with restrictions, as explained in the next section.  
The switch can forward healthcheck messages over the control VLAN  
from the master node and respond appropriately when notified of a ring  
fault by the master node.  
The master node generates a variety of messages over the control VLAN  
for monitoring the health of the ring and notifying the nodes of changes to  
the ring’s status. Two of these messages are the Ring-Down-Flush-FDB  
and Ring-Up-Flush-FDB messages. The first message notifies the nodes  
of a ring fault condition and the second signals the reestablishment of the  
ring.  
The AT-9400 Switch and EPSR snooping react to these messages by  
flushing the addresses learned on the two ring ports of the control VLAN  
from the forwarding database, so that the switch can relearn the  
addresses. These are the only two EPSR messages that EPSR snooping  
can react to. It should be noted that EPSR snooping cannot generate any  
EPSR messages itself.  
To configure the AT-9400 Switch as a transit node you need to create the  
control and data VLANs of the individual ring domains. As explained in the  
EPSR chapter in the AlliedWare OS Software Reference Guide, several  
domains can share the same physical network, but they must operate as  
logically separate VLAN groups. For information on VLANs, refer to  
Section III: Snooping Protocols  
189  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 18: Ethernet Protection Switching Ring Snooping  
After creating the VLANs, you activate EPSR snooping by specifying the  
control VLAN with the ENABLE EPSRSNOOPING command. The switch  
immediately begins to monitor the VLAN for control messages from the  
master switch and reacts accordingly should it receive EPSR messages  
on one of the two ports of the VLAN.  
190  
Section III: Snooping Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Restrictions  
EPSR snooping has three important restrictions. All the restrictions are  
related to control EPSR messages and the fact that EPSR snooping can  
not generate these messages.  
The AT-9400 Switch cannot fulfill the role of master node of a ring because  
EPSR snooping does not generate EPSR control messages. That function  
must be assigned to another Allied Telesis switch that supports EPSR,  
such as the AT-8948 Fast Ethernet Layer 3 Switch. (For a list of Allied  
Telesis products that support EPSR, refer to the company’s web site or  
contact your sales representative.)  
The second restriction is EPSR snooping does not support the transit node  
unsolicited method of fault detection. When a break occurs in a ring, the  
transit nodes on either side of the break can notify the master node by  
sending a “links down” message over the control VLAN of the ring. This  
method of fault detection and notification can be a faster way for the  
master node to become aware of a problem than with the healthcheck  
message. However, since EPSR snooping can not generate the “links  
down” message, the AT-9400 Switch can not initiate this type of fault  
notification.  
The final restriction of EPSR snooping concerns how the switch responds  
in the unlikely event it becomes isolated by disruptions in the ring on either  
side of it. Because the switch can not generate the EPSR “links up”  
message, a failure of a ring at two or more places may leave the AT-9400  
Switch and EPSR snooping isolated until all of the breaks have been  
repaired and the entirety of the ring is restored.  
The problem is illustrated in Figure 17. The example is a ring of four  
nodes: a master switch, two transit nodes, and the AT-9400 Switch  
running EPSR snooping. The ring has a control VLAN along with one or  
more data VLANs. The AT-9400 Switch is isolated from the ring because  
of two breaks, one between it and the master node and another between it  
and a transit node.  
Section III: Snooping Protocols  
191  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 18: Ethernet Protection Switching Ring Snooping  
AT-8948 Switch  
Master Node  
S
P
Transit Node  
AT-9400 Switch  
Transit Node  
Figure 17. Double Fault Condition in EPSR Snooping  
Now assume the link is reestablished between the switch and transit node.  
At that point, the port on the transit node enters a preforwarding state in  
which it forwards EPSR packets over the control VLAN to the AT-9400  
Switch. However, the transit node does not forward traffic over the data  
VLANs of the ring until it receives a “links up” message from the unit on  
the other side of the repaired break, in this case the AT-9400 Switch.  
Since EPSR snooping is not capable of generating EPSR messages, the  
transit node does not receive the anticipated signal, and so the data  
VLANs remain inactive. As a consequence, the AT-9400 Switch remains  
isolated from the ring until the other break is repaired and the master  
switch sends the EPSR Ring-Up-Flush-FDB message, which, in effect,  
initializes the ring.  
192  
Section III: Snooping Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Guidelines  
The guidelines to EPSR snooping are:  
ˆ The AT-9400 Switch can support up to sixteen control VLANs and so  
up to sixteen EPSR instances.  
ˆ The AT-9400 Switch can not be the master node of a ring.  
ˆ EPSR snooping does not support the transit node unsolicited method  
of fault notification.  
ˆ The switch must be operating in the user-configure VLAN mode to  
support the feature. EPSR snooping is not supported in the Multiple  
VLAN mode or the 802.1Q-compliant Multiple VLAN mode.  
ˆ The control VLAN must have exactly two ports. The only exception to  
this rule is if the ports of the control VLAN are part of a static port trunk.  
The ports, which must be tagged members of the VLAN, are used as  
the ring’s ports of the EPSR instance.  
ˆ The ports of the control VLAN and the data VLANs of an EPSR  
instance cannot be running LACP, STP, or GARP.  
ˆ The control VLAN cannot be part of another EPSR instance as either a  
control or data VLAN.  
ˆ EPSR snooping is only compatible with the EPSR feature on Allied  
Telesis products.  
Section III: Snooping Protocols  
193  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 18: Ethernet Protection Switching Ring Snooping  
194  
Section III: Snooping Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Section IV  
SNMPv3  
The chapter in this section contains overview information on SNMPv3. The  
chapter is:  
Section IV: SNMPv3  
195  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
196  
Section IV: SNMPv3  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 19  
SNMPv3  
This chapter provides a description of the AT-S63 implementation of the  
SNMPv3 protocol. The following sections are provided:  
Section IV: SNMPv3  
197  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 19: SNMPv3  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
198  
Section IV: SNMPv3  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
The SNMPv3 protocol builds on the existing SNMPv1 and SNMPv2c  
protocol implementation which is described in Chapter 3, “SNMPv1 and  
SNMPv2c” on page 65. In SNMPv3, User-based Security Model (USM)  
authentication is implemented along with encryption, allowing you to  
configure a secure SNMP environment.  
In addition, SNMP terminology changes in the SNMPv3 protocol. In the  
SNMPv1 and SNMPv2c protocols, the terms agent and manager are  
used. An agent is an SNMP user while a manager is an SNMP host. In the  
SNMPv3 protocol, agents and managers are called entities. In any  
SNMPv3 communication, there is an authoritative entity and a non-  
authoritative entity. The authoritative entity checks the authenticity of the  
non-authoritative entity. And, the non-authoritative entity checks the  
authenticity of the authoritative entity.  
With the SNMPv3 protocol, you create users, determine the protocol used  
for message authentication as well as determine if data transmitted  
between two SNMP entities is encrypted. In addition, you can restrict user  
privileges by determining the user’s view of the Management Information  
Bases (MIB). In this way, you restrict which MIBs the user can display and  
modify. In addition, you can restrict the types of messages, or traps, the  
user can send. (A trap is a type of SNMP message.)  
After you have created a user, you define SNMPv3 message notification.  
This consists of determining where messages are sent and what types of  
messages can be sent. This configuration is similar to the SNMPv1 and  
SNMPv2c configuration because you configure IP addresses of trap  
receivers, or hosts. In addition, with the SNMPv3 implementation you  
decide what types of messages are sent.  
Note  
For the SNMP RFCs supported by this release of the AT-S63  
This section further describes the features of the SNMPv3 protocol. The  
following subsections are included:  
Section IV: SNMPv3  
199  
Download from Www.Somanuals.com. All Manuals Search And Download.  
               
Chapter 19: SNMPv3  
SNMPv3 Authentication Protocols  
The SNMPv3 protocol supports two authentication protocols—HMAC-  
MD5-96 (MD5) and HMAC-SHA-96 (SHA). Both MD5 and SHA use an  
algorithm to generate a message digest. Each authentication protocol  
authenticates a user by checking the message digest. In addition, both  
protocols use keys to perform authentication. The keys for both protocols  
are generated locally using the Engine ID, a unique identifier that is  
assigned to the switch automatically, and the user password. You modify  
a key only by modifying the user password.  
In addition, you have the option of assigning no user authentication. In this  
case, no authentication is performed for this user. You may want to make  
this configuration for someone with super-user capabilities.  
Note  
The keys generated by the MD5 and SHA protocols are specific to  
the SNMPv3 protocol. They have no relation to the SSL and SSH  
keys for encryption.  
200  
Section IV: SNMPv3  
Download from Www.Somanuals.com. All Manuals Search And Download.  
         
AT-S63 Management Software Features Guide  
SNMPv3 Privacy Protocol  
After you have configured an authentication protocol, you have the option  
of assigning a privacy protocol if you have the encrypted version of the  
AT-S63 software. In SNMPv3 protocol terminology, privacy is equivalent to  
encryption. Currently, the DES protocol is the only encryption protocol  
supported. The DES privacy protocol requires the authentication protocol  
to be configured as either MD5 or SHA.  
If you assign a DES privacy protocol to a user, then you are also required  
to assign a privacy password. If you choose to not assign a privacy value,  
then SNMPv3 messages are sent in plain text format.  
Section IV: SNMPv3  
201  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 19: SNMPv3  
SNMPv3 MIB Views  
The SNMPv3 protocol allows you to configure MIB views for users and  
groups. The MIB tree is defined by RFC 1155 (Structure of Management  
root  
ccitt (0)  
iso (1)  
joint-iso-ccitt (2)  
member-body (2) identified-organization (3)  
standard (0)  
registration-authority (1)  
dod (6)  
internet (1)  
directory (1)  
mgmt (2)  
mib-2 (1)  
experimental (3)  
private (4)  
system (1)  
at (3)  
icmp (5)  
udp (7)  
cmot (9)  
snmp (11)  
host (25)  
interfaces (2)  
ip (4)  
tcp (6)  
egp (8)  
transmission (10)  
dot1 dBridge (117)  
Figure 18. MIB Tree  
The AT-S63 software supports the MIB tree, starting with the Internet  
MIBs, as defined by 1.3.6.1. There are two ways to specify a MIB view.  
You can enter the OID number of the MIB view or its equivalent text name.  
For example, to specify MIBs in the Internet view, you can enter the OID  
format “1.3.6.1” or the text name “internet.”  
In addition, you can define a MIB view that the user can access or a MIB  
view that the user cannot access. When you want to permit a user to  
access a MIB view, you include a particular view. When you want to deny  
a user access to a MIB view, you exclude a particular view.  
202  
Section IV: SNMPv3  
Download from Www.Somanuals.com. All Manuals Search And Download.  
           
AT-S63 Management Software Features Guide  
After you specify a MIB subtree view you have the option of further  
restricting a view by defining a subtree mask. The relationship between a  
MIB subtree view and a subtree mask is analogous to the relationship  
between an IP address and a subnet mask. The switch uses the subnet  
mask to determine which portion of an IP address represents the network  
address and which portion represents the node address. In a similar way,  
the subtree mask further refines the subtree view and enables you to  
restrict a MIB view to a specific row of the OID MIB table. You need a  
thorough understanding of the OID MIB table to define a subtree mask.  
Section IV: SNMPv3  
203  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 19: SNMPv3  
SNMPv3 Storage Types  
Each SNMPv3 table entry has its own storage type. You can choose  
between nonvolatile storage which allows you to save the table entry or  
volatile storage which does not allow you to save an entry. If you select the  
volatile storage type, when you power off the switch your SNMPv3  
configuration is lost and cannot be recovered.  
At each SNMPv3 menu, you are prompted to configure a storage type.  
You do not have to configure the same storage type value for each table  
entry.  
204  
Section IV: SNMPv3  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
AT-S63 Management Software Features Guide  
SNMPv3 Message Notification  
When you generate an SNMPv3 message from the switch, there are three  
basic pieces of information included in the message:  
ˆ The type of message  
ˆ The destination of the message  
ˆ SNMP security information  
To configure the type of message, you need to define if you are sending a  
Trap or Inform message. Basically, the switch expects a response to an  
Inform message and the switch does not expect a response to a Trap  
message. These two message types are defined in the SNMPv3 (RFC  
2571-6).  
To determine the destination of the message, you configure the IP address  
of the host. This configuration is similar to the SNMPv1 and SNMPv2c  
configuration.  
The SNMP security information consists of information about the following:  
ˆ User  
ˆ View of the MIB Tree  
ˆ Security Level  
ˆ Security Model  
ˆ Authentication Level  
ˆ Privacy Protocol  
ˆ Group  
To configure the SNMP security information, you associate a user and its  
related information—View, Security Level, Security Model, Authentication  
Level, Privacy Protocol and Group—with the type of message and the host  
IP address.  
Section IV: SNMPv3  
205  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 19: SNMPv3  
SNMPv3 Tables  
The SNMPv3 configuration is neatly divided into configuring SNMPv3 user  
information and configuring the message notification. You must configure  
all seven tables to successfully configure the SNMPv3 protocol. You use  
the following tables for user configuration:  
ˆ Configure SNMPv3 User Table  
ˆ Configure SNMPv3 View Table  
ˆ Configure SNMPv3 Access Table  
ˆ Configure SNMPv3 SecurityToGroup Table  
First, you create a user in the Configure SNMPv3 User Table. Then you  
define the MIB view this user has access to in the Configure SNMPv3  
View Table. To configure a security group and associate a MIB view to a  
security group, you configure the Configure SNMPv3 Access Table.  
Finally, configure the Configure SNMPv3 SecurityToGroup menu to  
associate a user to a security group. See Figure 19 for an illustration of  
how the user configuration tables are linked.  
SNMPv3 User Table  
SNMPv3 View Table  
Linked by User  
Linked by View Name  
Name/Security  
Name  
SNMPv3 Access Table  
Linked by Group Name  
SNMPv3 Security To Group Table  
Figure 19. SNMPv3 User Configuration Process  
In general, you focus on configuring security groups and then add and  
delete users from the groups as needed. For example, you may want to  
have two groups—one for manager privileges and a second one for  
example of manager and operator configurations.  
After you configure an SNMPv3 user, you need to configure SNMPv3  
message notification. This configuration is accomplished with the following  
tables:  
206  
Section IV: SNMPv3  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
ˆ Configure SNMPv3 Notify Table  
ˆ Configure SNMPv3 Target Address Table  
ˆ Configure SNMPv3 Target Parameters Table  
You start the message notification configuration by defining the type of  
message you want to send with the SNMPv3 Notify Table. Then you  
define a IP address that is used for notification in the Configure SNMPv3  
Target Address Table. This is the IP address of the SNMPv3 host. Finally,  
you associate the trap information with a user by configuring the Configure  
SNMPv3 Target Parameters Table.  
See Figure 20 for an illustration of how the message notification tables are  
linked.  
SNMPv3 Notify Table  
Linked by Notify Tag  
SNMPv3 Target Address Table  
SNMPv3 Target Parameter Table  
Linked by Target Parameter Name  
Linked by User Name  
or Security Name  
SNMPv3 User Table  
SNMPv3 View Table  
SNMPv3 Access Table  
Linked by  
Security Name  
and  
Security Model  
Linked by View Name  
Linked by Group Name  
SNMPv3 SecurityToGroup Table  
Figure 20. SNMPv3 Message Notification Process  
For a more detailed description of the SNMPv3 Tables, see the following  
subsections:  
Section IV: SNMPv3  
207  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 19: SNMPv3  
SNMPv3 User The Configure SNMPv3 User Table menu allows you to create an  
SNMPv3 user and provides the options of configuring authentication and  
Table  
privacy protocols. With the SNMPv3 protocol, users are authenticated  
when they send and receive messages. In addition, you can configure a  
privacy protocol and password so messages a user sends and receives  
are encrypted. The DES privacy algorithm uses the privacy password and  
the Engine ID to generate a key that is used for encryption. Lastly, you can  
configure a storage type for this table entry which allows you to save this  
user and its related configuration to flash memory.  
SNMPv3 View The Configure SNMPv3 View Table menu allows you to create a view of  
the MIB OID Table. First, you configure a view of a subtree. Then you  
Table  
have the option of configuring a Subtree Mask that further refines the  
subtree view. For example, you can use a Subtree Mask to restrict a  
user’s view to one row of the MIB OID Table. In addition, you can chose to  
include or exclude a view. As a result, you can let a user see a particular  
view or prevent a user from seeing a particular view. Lastly, you can  
configure a storage type for this table entry which allows you to save this  
view to flash memory.  
SNMPv3 Access The Configure SNMPv3 Access Table menu allows you to configure a  
security group. After you create a security group, you assign a set of users  
Table  
with the same access privileges to this group using the SNMPv3  
SecurityToGroup Table. Consider the types of groups you want to create  
and the types of access privileges each group will have. In this way, you  
can more easily keep track of your users as belonging to one or two  
groups.  
For each group, you can assign read, write, and notify views of the MIB  
table. The views you assign here have been previously defined in the  
Configure SNMPv3 View Table menu. For example, the Read View allows  
group members to view the specified portion of the OID MIB table. The  
Write View allows group members to write to, or modify, the MIBs in the  
specified MIB view. The Notify View allows group members to send trap  
messages defined by the MIB view. Lastly, you can configure a storage  
type for this table entry which allows you to save this view to flash  
memory.  
SNMPv3 The Configure SNMPv3 SecurityToGroup Table menu allows you to  
associate a User Name with a security group called a Group Name. The  
User Name is previously configured with the Configure SNMPv3 User  
Table menu. The security group is previously configured with the  
SecurityToGroup  
Table  
Configure SNMPv3 Access Table menu. Lastly, you can configure a  
storage type for this table entry which allows you to save the entry to flash  
memory.  
208  
Section IV: SNMPv3  
Download from Www.Somanuals.com. All Manuals Search And Download.  
               
AT-S63 Management Software Features Guide  
SNMPv3 Notify The Configure SNMPv3 Notify Table menu allows you to define the type of  
message that is sent from the switch to the SNMP host. In addition, you  
Table  
have the option of defining the message type as either an Inform or a Trap  
message. The difference between these two types of messages is that  
when a switch sends an Inform message, the switch expects a response  
from the host. In comparison, the switch does not expect the host to  
respond to Trap messages.  
In addition, you define a Notify Tag that links an SNMPv3 Notify Table  
entry to the host IP address defined in the Configure SNMPv3 Target  
Address Table menu. Lastly, you can configure a storage type for this  
table entry which allows you to save the entry to flash memory.  
SNMPv3 Target The Configure SNMPv3 Target Address Table menu allows you to  
configure the IP address of the host. Also, in an SNMPv3 Target Address  
Address Table  
Table entry, you configure the values of the Tag List parameter with the  
previously defined Notify Tag parameter values. The Notify Tag parameter  
is configured in the Configure SNMPv3 Notify Table. In this way, the Notify  
and Target Address tables are linked. Lastly, you can configure a storage  
type for this table entry which allows you to save the entry to flash  
memory.  
SNMPv3 Target The Configure SNMPv3 Target Parameters Table menu allows you to  
define which user can send messages to the host IP address defined in  
Parameters Table  
the Configure SNMPv3 Target Address Table. The user and its associated  
information is previously configured in the Configure SNMPv3 User Table,  
SNMPv3 View Table, SNMPv3 Access Table, and SNMPv3  
SecurityToGroup Table. Lastly, you can configure a storage type for this  
table entry which allows you to save the entry to flash memory.  
SNMPv3 The Configure SNMPv3 Community Table menu allows you to configure  
SNMPv1 and SNMPv2c communities. If you are going to use the SNMPv3  
Tables to configure SNMPv1 and SNMPv2c communities, start with the  
SNMPv3 Community Table.  
Community  
Table  
Section IV: SNMPv3  
209  
Download from Www.Somanuals.com. All Manuals Search And Download.  
               
Chapter 19: SNMPv3  
SNMPv3 Configuration Example  
You may want to have two classes of SNMPv3 users—Managers and  
Operators. In this scenario, you would configure one group, called  
Managers, with full access privileges. Then you would configure a second  
group, called Operators, with monitoring privileges only. For a detailed  
example of this configuration, see Appendix B, “SNMPv3 Configuration  
210  
Section IV: SNMPv3  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Section V  
Spanning Tree Protocols  
The section has the following chapters:  
Section V: Spanning Tree Protocols  
211  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
212  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 20  
Spanning Tree and Rapid Spanning Tree  
Protocols  
This chapter provides background information on the Spanning Tree  
Protocol (STP) and Rapid Spanning Tree Protocol (RSTP). The sections in  
this chapter include:  
Note  
For detailed information on the Spanning Tree Protocol, refer to  
IEEE Std 802.1D. For detailed information on the Rapid Spanning  
Tree Protocol, refer to IEEE Std 802.1w.  
Section V: Spanning Tree Protocols  
213  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 20: Spanning Tree and Rapid Spanning Tree Protocols  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
214  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
The performance of a Ethernet network can be negatively impacted by the  
formation of a data loop in the network topology. A data loop exists when  
two or more nodes on a network can transmit data to each other over more  
than one data path. The problem that data loops pose is that data packets  
can become caught in repeating cycles, referred to as broadcast storms,  
that needlessly consume network bandwidth and can significantly reduce  
network performance.  
STP and RSTP prevent data loops from forming by ensuring that only one  
path exists between the end nodes in your network. Where multiple paths  
exist, these protocols place the extra paths in a standby or blocking mode,  
leaving only one main active path.  
STP and RSTP can also activate a redundant path if the main path goes  
down. So not only do these protocols guard against multiple links between  
segments and the risk of broadcast storms, but they can also maintain  
network connectivity by activating a backup redundant path in case a main  
link fails.  
Where the two protocols differ is in the time each takes to complete the  
process referred to as convergence. When a change is made to the  
network topology, such as the addition of a new bridge, a spanning tree  
protocol must determine whether there are redundant paths that must be  
blocked to prevent data loops, or activated to maintain communications  
between the various network segments. This is the process of  
convergence.  
With STP, convergence can take up to a minute to complete in a large  
network. This can result in the loss of communication between various  
parts of the network during the convergence process, and the subsequent  
lost of data packets.  
RSTP is much faster. It can complete a convergence in seconds, and so  
greatly diminish the possible impact the process can have on your  
network.  
Only one spanning tree can be active on the switch at a time. The default  
is RSTP.  
The STP implementation on the AT-S63 Management Software complies  
with the IEEE 802.1d standard. The RSTP implementation complies with  
the IEEE 802.1w standard. The following subsections provide a basic  
overview on how STP and RSTP operate and define the different  
parameters that you can adjust.  
Section V: Spanning Tree Protocols  
215  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 20: Spanning Tree and Rapid Spanning Tree Protocols  
Bridge Priority and the Root Bridge  
The first task that bridges perform when a spanning tree protocol is  
activated on a network is the selection of a root bridge. A root bridge  
distributes network topology information to the other network bridges and  
is used by the other bridges to determine if there are redundant paths in  
the network.  
A root bridge is selected by the bridge priority number, also referred to as  
the bridge identifier, and sometimes the bridge’s MAC address. The bridge  
with the lowest bridge priority number in the network is selected as the  
root bridge. If two or more bridges have the same bridge priority number,  
of those bridges the one with the lowest MAC address is designated as  
the root bridge.  
You can change the bridge priority number in the AT-S63 Management  
Software. You can designate which switch on your network you want as  
the root bridge by giving it the lowest bridge priority number. You might  
also consider which bridge should function as the backup root bridge in  
the event you need to take the primary root bridge offline, and assign that  
bridge the second lowest bridge identifier number.  
The bridge priority has a range 0 to 61440 in increments of 4096. To make  
this easier for you, the AT-S63 Management Software divides the range  
into increments. You specify the increment that represents the desired  
bridge priority value. The range is divided into sixteen increments, as  
shown in Table 14.  
Table 14. Bridge Priority Value Increments  
Bridge  
Priority  
Bridge  
Priority  
Increment  
Increment  
0
0
8
32768  
1
2
3
4
5
6
7
4096  
9
36864  
40960  
45056  
49152  
53248  
57344  
61440  
8192  
10  
11  
12  
13  
14  
15  
12288  
16384  
20480  
24576  
28672  
216  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Path Costs and After the root bridge has been selected, the bridges determine if the  
network contains redundant paths and, if one is found, select a preferred  
path while placing the redundant paths in a backup or blocking state.  
Port Costs  
Where there is only one path between a bridge and the root bridge, the  
bridge is referred to as the designated bridge and the port through which  
the bridge is communicating with the root bridge is referred to as the root  
port.  
If redundant paths exist, the bridges that are a part of the paths must  
determine which path will be the primary, active path, and which path(s)  
will be placed in the standby, blocking mode. This is accomplished by an  
determination of path costs. The path offering the lowest cost to the root  
bridge becomes the primary path and all other redundant paths are placed  
into blocking state.  
Path cost is determined by evaluating port costs. Every port on a bridge  
participating in STP has a cost associated with it. The cost of a port on a  
bridge is typically based on port speed. The faster the port, the lower the  
port cost. The exception to this is the ports on the root bridge, where all  
ports have a port cost of 0.  
Path cost is simply the sum of the port costs between a bridge and the root  
bridge.  
The port cost of a port on the AT-9400 Switch is adjustable through the  
AT-S63 Management Software. For STP, the range is 0 to 65,535. For  
RSTP, the range is 0 to 20,000,000.  
Port cost also has an Auto-Detect feature. This feature allows spanning  
tree to automatically set the port cost according to the speed of the port,  
assigning a lower value for higher speeds. Auto-Detect is the default  
setting. Table 17 lists the STP port costs with Auto-Detect.  
Table 15. STP Auto-Detect Port Costs  
Port Speed  
10 Mbps  
Port Cost  
100  
100 Mbps  
10  
4
1000 Mbps  
Section V: Spanning Tree Protocols  
217  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 20: Spanning Tree and Rapid Spanning Tree Protocols  
Table 16 lists the STP port costs with Auto-Detect when a port is part of a  
port trunk.  
Table 16. STP Auto-Detect Port Trunk Costs  
Port Speed  
Port Cost  
10 Mbps  
4
4
2
100 Mbps  
1000 Mbps  
Table 17 lists the RSTP port costs with Auto-Detect.  
Table 17. RSTP Auto-Detect Port Costs  
Port Speed  
Port Cost  
10 Mbps  
2,000,000  
200,000  
20,000  
100 Mbps  
1000 Mbps  
Table 18 lists the RSTP port costs with Auto-Detect when the port is part  
of a port trunk.  
Table 18. RSTP Auto-Detect Port Trunk Costs  
Port Speed  
Port Cost  
20,000  
10 Mbps  
100 Mbps  
1000 Mbps  
20,000  
2,000  
You can override Auto-Detect and set the port cost manually.  
Port Priority If two paths have the same port cost, the bridges must select a preferred  
path. In some instances this can involve the use of the port priority  
parameter. This parameter is used as a tie breaker when two paths have  
the same cost.  
The range for port priority is 0 to 240. As with bridge priority, this range is  
broken into increments, in this case multiples of 16. To select a port  
priority for a port, you enter the increment of the desired value. Table 19  
lists the values and increments. The default value is 128, which is  
increment 8.  
218  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
         
AT-S63 Management Software Features Guide  
Table 19. Port Priority Value Increments  
Bridge  
Priority  
Bridge  
Priority  
Increment  
Increment  
0
0
8
128  
1
2
3
4
5
6
7
16  
32  
48  
64  
80  
96  
112  
9
144  
160  
176  
192  
208  
224  
240  
10  
11  
12  
13  
14  
15  
Section V: Spanning Tree Protocols  
219  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 20: Spanning Tree and Rapid Spanning Tree Protocols  
Forwarding Delay and Topology Changes  
If there is a change in the network topology due to a failure, removal, or  
addition of any active components, the active topology also changes. This  
may trigger a change in the state of some blocked ports. However, a  
change in a port state is not activated immediately.  
It might take time for the root bridge to notify all bridges that a topology  
change has occurred, especially if it is a large network. If a topology  
change is made before all bridges have been notified, a temporary data  
loop could occur, and that could adversely impact network performance.  
To forestall the formation of temporary data loops during topology  
changes, a port designated to change from blocking to forwarding passes  
through two additional states—listening and learning—before it begins to  
forward frames. The amount of time a port spends in these states is set by  
the forwarding delay value. This value states the amount of time that a  
port spends in the listening and learning states prior to changing to the  
forwarding state.  
The forwarding delay value is adjustable in the AT-S63 Management  
Software. The appropriate value for this parameter depends on a number  
of variables; the size of your network is a primary factor. For large  
networks, you should specify a value large enough to allow the root bridge  
sufficient time to propagate a topology change throughout the entire  
network. For small networks, you should not specify a value so large that a  
topology change is unnecessarily delayed, which could result in the delay  
or loss of some data packets.  
Note  
The forwarding delay parameter applies only to ports on the switch  
that are operating STP-compatible mode.  
Hello Time and The bridges that are part of a spanning tree domain communicate with  
each other using a bridge broadcast frame that contains a special section  
devoted to carrying STP or RSTP information. This portion of the frame is  
referred to as the bridge protocol data unit (BPDU). When a bridge is  
Bridge Protocol  
Data Units  
(BPDU)  
brought online, it issues a BPDU in order to determine whether a root  
bridge has already been selected on the network, and if not, whether it has  
the lowest bridge priority number of all the bridges and should therefore  
become the root bridge.  
The root bridge periodically transmits a BPDU to determine whether there  
have been any changes to the network topology and to inform other  
bridges of topology changes. The frequency with which the root bridge  
sends out a BPDU is called the hello time. This is a value that you can set  
in the AT-S63 Management Software. The interval is measured in  
220  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
AT-S63 Management Software Features Guide  
seconds and the default is two seconds. Consequently, if the AT-9400  
Switch is selected as the root bridge of a spanning tree domain, it  
transmits a BPDU every two seconds.  
Point-to-Point  
and Edge Ports  
Note  
This section applies only to RSTP.  
Part of the task of configuring RSTP is defining the port types on the  
bridge. This relates to the device(s) connected to the port. With the port  
types defined, RSTP can reconfigure a network much quicker than STP  
when a change in network topology is detected.  
There are two possible selections:  
ˆ Point-to-point port  
ˆ Edge port  
If a bridge port is operating in full-duplex mode, than the port is functioning  
as a point-to-point port. Figure 21 illustrates two AT-9400 Switches that  
are connected with one data link. With the link operating in full-duplex, the  
ports are point-to-point ports.  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
CLASS  
1
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Point-to-Point Ports  
(Full-duplex Mode)  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
10/100 LINK  
L/A  
D/C  
1000 LINK  
FDX  
/
ACT  
/
ACT  
CLASS  
1
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
SFP  
23  
SFP  
24  
1000 LINK  
/
ACT  
L/A  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Figure 21. Point-to-Point Ports  
If a port is operating in half-duplex mode and is not connected to any  
further bridges participating in STP or RSTP, then the port is an edge port.  
Figure 22 illustrates an edge port on an AT-9400 Switch. The port is  
connected to an Ethernet hub, which in turn is connected to a series of  
Ethernet workstations. This is an edge port because it is connected to a  
device operating at half-duplex mode and there are no participating STP or  
RSTP devices connected to it.  
Section V: Spanning Tree Protocols  
221  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 20: Spanning Tree and Rapid Spanning Tree Protocols  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
CLASS  
1
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
8  
20  
22  
24R  
Edge Port  
1
5
2
3
7
4
8
6
1
2
3
4
5
6
7
8
Figure 22. Edge Port  
A port can be both a point-to-point and an edge port at the same time. It  
operates in full-duplex and has no STP or RSTP devices connected to it.  
Figure 23 illustrates a port functioning as both a point-to-point and edge  
port.  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
CLASS  
1
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
8
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
23  
24  
POWER  
2
4
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Point-to-Point and Edge Port  
Workstation  
(Full-duplex Mode)  
Figure 23. Point-to-Point and Edge Port  
Determining whether a bridge port is point-to-point, edge, or both, can be  
a bit confusing. For that reason, do not change the default values for this  
RSTP feature unless you have a good grasp of the concept. In most  
cases, the default values work well.  
222  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Mixed STP and RSTP Networks  
RSTP IEEE 802.1w is fully compliant with STP IEEE 802.1d. Your network  
can consist of bridges running both protocols. STP and RSTP in the same  
network can operate together to create a single spanning tree domain.  
If you decide to activate spanning tree on the switch, there is no reason  
not to activate RSTP on the AT-9400 Switch even when all other switches  
are running STP. The switch can combine its RSTP with the STP of the  
other switches. The switch monitors the traffic on each port for BPDU  
packets. Ports that receive RSTP BPDU packets operates in RSTP mode  
while ports receiving STP BPDU packets operate in STP mode.  
Section V: Spanning Tree Protocols  
223  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 20: Spanning Tree and Rapid Spanning Tree Protocols  
Spanning Tree and VLANs  
The spanning tree implementation in the AT-S63 Management Software is  
a single-instance spanning tree. The switch supports just one spanning  
tree. You cannot define multiple spanning trees.  
The single spanning tree encompasses all ports on the switch. If the ports  
are divided into different VLANs, the spanning tree crosses the VLAN  
boundaries. This point can pose a problem in networks containing multiple  
VLANs that span different switches and are connected with untagged  
ports. In this situation, STP blocks a data link because it detects a data  
loop. This can cause fragmentation of your VLANs.  
This issue is illustrated in Figure 24. Two VLANs, Sales and Production,  
span two AT-9400 Switches. Two links consisting of untagged ports  
connect the separate parts of each VLAN. If STP or RSTP is activated on  
the switches, one of the links is disabled. In the example, the port on the  
top switch that links the two parts of the Production VLAN is changed to  
the block state. This leaves the two parts of the Production VLAN unable  
to communicate with each other.  
Sales  
VLAN  
Production  
VLAN  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
CLASS  
1
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Blocked Port  
Blocked Data Link  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
10/100 LINK  
L/A  
D/C  
1000 LINK  
FDX  
/
ACT  
/
ACT  
CLASS  
1
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
2
3
4
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
SFP  
23  
SFP  
24  
1000 LINK  
/
ACT  
L/A  
23  
24  
POWER  
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Sales  
VLAN  
Production  
VLAN  
Figure 24. VLAN Fragmentation  
You can avoid this problem by not activating spanning tree or by  
connecting VLANs using tagged instead of untagged ports. (For  
information on tagged and untagged ports, refer to Chapter 22, “Port-  
224  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 21  
Multiple Spanning Tree Protocol  
This chapter provides background information on the Multiple Spanning  
Tree Protocol (MSTP). The sections in this chapter include:  
Section V: Spanning Tree Protocols  
225  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 21: Multiple Spanning Tree Protocol  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
226  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Overview  
Protocols” on page 213, STP and RSTP are referred to as single-instance  
spanning trees that search for physical loops across all VLANs in a  
bridged network. When loops are detected, the protocols stop the loops by  
placing one or more bridge ports in a blocking state.  
can result in VLAN fragmentation where VLANs that span multiple bridges  
are connected together with untagged ports. The untagged ports creating  
the links can represent a physical loop in the network, which are blocked  
by spanning tree. This can result in a loss of communication between  
different parts of the same VLAN.  
One way to resolve this, other than by not activating spanning tree on your  
network, is to link the switches using tagged ports, which can handle traffic  
from multiple VLANs simultaneously. The drawback to this approach is  
that the link formed by the tagged ports can create a bottleneck to your  
Ethernet traffic, resulting in reduced network performance.  
Another approach is to use the Multiple Spanning Tree Protocol (MSTP).  
This spanning tree shares many of the same characteristics as RSTP. It  
features rapid convergence and has many of the same parameters. But  
the main difference is that while RSTP, just like STP, supports only a  
single-instance spanning tree, MSTP supports multiple spanning trees  
within a network.  
The following sections describe some of the terms and concepts relating to  
MSTP. If you are not familiar with spanning tree or RSTP, you should first  
Note  
Do not activate MSTP on the AT-9400 Switch without first  
familiarizing yourself with the following concepts and guidelines.  
Unlike STP and RSTP, you cannot activate this spanning tree  
protocol on a switch without first configuring the protocol  
parameters.  
Note  
The AT-S63 MSTP implementation complies fully with the new IEEE  
802.1s standard and should be interoperable with any other  
vendor’s fully compliant 802.1s implementation.  
Section V: Spanning Tree Protocols  
227  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 21: Multiple Spanning Tree Protocol  
Multiple Spanning Tree Instance (MSTI)  
The individual spanning trees in MSTP are referred to as Multiple  
Spanning Tree Instances (MSTIs). A MSTI can span any number of  
AT-9400 Switches. The switch can support up to 16 MSTIs at a time.  
To create a MSTI, you first assign it a number, referred to as the MSTI ID.  
The range is 1 to 15. (The switch is shipped with a default MSTI with an  
MSTI ID of 0. This default spanning tree instance is discussed later in  
After you have selected an MSTI ID, you need to define the scope of the  
MSTI by assigning one or more VLANs to it. An instance can contain any  
number of VLANs, but a VLAN can belong to only one MSTI at a time.  
Following are several examples. Figure 25 illustrates two AT-9400  
Switches, each containing the two VLANs Sales and Production. The two  
parts of each VLAN are connected with a direct link using untagged ports  
on both switches. If the switches were running STP or RSTP, one of the  
links would be blocked because the links constitute a physical loop. Which  
link would be blocked depends on the STP or RSTP bridge settings. In  
Figure 25, the link between the two parts of the Production VLAN is  
blocked, resulting in a loss of communications between the two parts of  
the Production VLAN.  
228  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Sales  
VLAN  
Production  
VLAN  
1
3
5
7
9
11  
1
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
L/A  
D/C  
1000 LINK  
FDX  
/
ACT  
10/100 LINK  
HDX COL  
11 13 15 17 19 21 23R  
/
ACT  
CLASS  
1
LASER PRODUCT  
/
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
0  
22  
24R  
Blocked Port  
Untagged  
Ports  
Untagged  
Ports  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/GB Gigabit Ethernet Switch  
PORT ACTIVITY  
10/100 LINK  
L/A  
D/C  
1000 LINK  
FDX  
/
ACT  
/
ACT  
CLASS  
1
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
GBIC  
GBIC  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
GBIC  
1000 LINK  
/
ACT  
L/A  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
1
16  
18  
20  
22  
24R  
23  
24  
Figure 25. VLAN Fragmentation with STP or RSTP  
Section V: Spanning Tree Protocols  
229  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 21: Multiple Spanning Tree Protocol  
Figure 26 illustrates the same two AT-9400 Switches and the same two  
virtual LANs. But in this example, the two switches are running MSTP and  
the two VLANs have been assigned different spanning tree instances.  
Now that they reside in different MSTIs, both links remain active, enabling  
the VLANs to forward traffic over their respective direct link.  
Sales  
VLAN in  
MSTI 1  
Production  
VLAN in  
MSTI 2  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
CLASS  
1
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
2
22  
24R  
Untagged  
Ports  
Untagged  
Ports  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/GB Gigabit Ethernet Switch  
PORT ACTIVITY  
10/100 LINK  
L/A  
D/C  
1000 LINK  
FDX  
/
ACT  
/
ACT  
CLASS  
1
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
GBIC  
GBIC  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
GBIC  
1000 LINK  
/
ACT  
L/A  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
23  
24  
Figure 26. MSTP Example of Two Spanning Tree Instances  
230  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
A MSTI can contain more than one VLAN. This is illustrated in Figure 27  
where there are two AT-9400 Switches with four VLANs. There are two  
MSTIs, each containing two VLANs. MSTI 1 contains the Sales and  
Presales VLANs and MSTI 2 contains the Design and Engineering VLANs.  
MSTI 1  
MSTI 2  
1
3
5
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
CLASS  
1
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
2  
24R  
Tagged  
Ports  
Tagged  
Ports  
1
3
5
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/GB Gigabit Ethernet Switch  
PORT ACTIVITY  
10/100 LINK  
L/A  
D/C  
1000 LINK  
FDX  
/
ACT  
/
ACT  
CLASS  
1
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
GBIC  
GBIC  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
GBIC  
1000 LINK  
/
ACT  
L/A  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
4  
16  
18  
20  
22  
24R  
23  
24  
Engineering  
VLAN  
Presales  
VLAN  
Sales  
VLAN  
Design  
VLAN  
Figure 27. Multiple VLANs in a MSTI  
In this example, because an MSTI contains more than one VLAN, the links  
between the VLAN parts is made with tagged, not untagged, ports so that  
they can carry traffic from more than one virtual LAN. Referring again to  
Figure 27, the tagged link in MSTI 1 is carrying traffic for both the Presales  
and Sales VLANs while the tagged link in MSTI 2 is carrying traffic for the  
Design and Engineering VLANs.  
Section V: Spanning Tree Protocols  
231  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 21: Multiple Spanning Tree Protocol  
MSTI Guidelines  
Following are several guidelines to keep in mind about MSTIs:  
ˆ The AT-9400 Switch can support up to 16 spanning tree instances,  
including the CIST.  
ˆ A MSTI can contain any number of VLANs.  
ˆ A VLAN can belong to only one MSTI at a time.  
ˆ A switch port can belong to more than one spanning tree instance at a  
time by being an untagged and tagged member of VLANs belonging to  
different MSTI’s. This is possible because a port can be in different  
MSTP states for different MSTI’s simultaneously. For example, a port  
can be in the MSTP blocking state for one MSTI and the forwarding  
state for another spanning tree instance. For further information, refer  
ˆ A router or Layer 3 network device is required to forward traffic  
between different VLANs.  
232  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
VLAN and MSTI Associations  
Part of the task to configuring MSTP involves assigning VLANs to  
spanning tree instances. The mapping of VLANs to MSTIs is called  
associations. A VLAN, either port-based or tagged, can belong to only one  
instance at a time, but an instance can contain any number of VLANs.  
Section V: Spanning Tree Protocols  
233  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 21: Multiple Spanning Tree Protocol  
Ports in Multiple MSTIs  
A port can be a member of more than one MSTI at a time if it is a tagged  
member of one or more VLANs assigned to different MSTI’s. In this  
circumstance, a port might be have to operate in different spanning tree  
states simultaneously, depending on the requirements of the MSTIs. For  
example, a port that belongs to two different VLANs in two different MSTIs  
might operate in the forwarding state in one MSTI and the blocking state in  
the other.  
A port’s MSTI parameter settings are divided into two groups. The first  
group is referred to as generic parameters. These are set just once on a  
port and apply to all the MSTI’s where the port is a member. One of these  
parameters is the external path cost, which sets the operating cost of a  
port connected to a device outside its region. A port, even if it belongs to  
multiple MSTI’s, can have only one external path cost. Other generic  
parameters designate the port as an edge port or a point-to-point port.  
The second group of port parameters can be set differently for each MSTI  
where a port is a member. One parameter, the internal path cost, specifies  
the operating cost of a port when it is connected to a bridge in the same  
MSTP region. The other parameter in this group sets the port priority,  
which acts as a tie breaker when two or more ports have equal costs to a  
regional root bridge.  
234  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Multiple Spanning Tree Regions  
Another important concept of MSTP is regions. A MSTP region is defined  
as a group of bridges that share exactly the same MSTI characteristics.  
Those characteristics are:  
ˆ Configuration name  
ˆ Revision number  
ˆ VLANs  
ˆ VLAN to MSTI ID associations  
A configuration name is a name assigned to a region to identify it. You  
must assign each bridge in a region exactly the same name; even the  
same upper and lowercase lettering. Identifying the regions in your  
network is easier if you choose names that are characteristic of the  
functions of the nodes and bridges of the region. Examples are Sales  
Region and Engineering Region.  
The revision number is an arbitrary number assigned to a region. This  
number can be used to keep track of the revision level of a region’s  
configuration. For example, you might use this value to maintain the  
number of times you revise a particular MSTP region. It is not important  
that you maintain this number, only that each bridge in a region has the  
same number.  
The bridges of a particular region must also have the same VLANs. The  
names of the VLANs and the VIDs must be same on all bridges of a  
region.  
Finally, the VLANs in the bridges must be associated to the same MSTIs.  
If any of the above information is different on two bridges, MSTP does  
consider the bridges as residing in different regions.  
Section V: Spanning Tree Protocols  
235  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 21: Multiple Spanning Tree Protocol  
Figure 28 illustrates the concept of regions. It shows one MSTP region  
consisting of two AT-9400 Switches. Each switch in the region has the  
same configuration name and revision level. The switches also have the  
same five VLANs and the VLANs are associated with the same MSTIs.  
Configuration Name: Marketing  
Region  
Revision Level: 1  
VLAN to MSTI Associations:  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT  
-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
F
AUL  
T
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
MASTER  
RPS  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
MSTI ID 1  
VLAN: Sales (VID 2)  
VLAN: Presales (VID 3)  
MSTI ID 2  
VLAN: Accounting (VID 4)  
Configuration Name: Marketing  
Region  
Revision Level: 1  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT  
-9424T/GB Gigabit Ethernet Switch  
PORT ACTIVITY  
VLAN to MSTI Associations:  
L/A  
D/C  
1000 LINK  
FDX  
/
ACT  
10/100 LINK  
HDX COL  
11 13 15 17 19 21 23R  
/
ACT  
CLASS  
1
LASER PRODUCT  
/
TERMINAL  
PORT  
STATUS  
1
3
5
7
9
GBIC  
GBIC  
L/A  
D/C  
L/A  
D/C  
F
AUL  
T
GBIC  
1000 LINK  
/
ACT  
MASTER  
RPS  
L/A  
23  
24  
POWER  
2
4
6
8
10 12 14 16 18 20 22 24R  
2
2
MSTI ID 1  
VLAN: Sales (VID 2)  
VLAN: Presales (VID 3)  
MSTI ID 2  
VLAN: Accounting (VID 4)  
Figure 28. Multiple Spanning Tree Region  
The AT-9400 Switch determines regional boundaries by examining the  
MSTP BPDUs received on the ports. A port that receives a MSTP BPDU  
from another bridge with regional information different from its own is  
considered to be a boundary port and the bridge connected to the port as  
belonging to another region.  
236  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
The same is true for any ports connected to bridges running the single-  
instance spanning tree STP or RSTP. Those ports are also considered as  
part of another region.  
Each MSTI functions as an independent spanning tree within a region.  
Consequently, each MSTI must have a root bridge to locate physical loops  
within the spanning tree instance. An MSTI’s root bridge is called a  
regional root. The MSTIs within a region may share the same regional root  
or they can have different regional roots.  
A regional root for an MSTI must be within the region where the MSTI is  
located. An MSTI cannot have a regional root that is outside its region.  
A regional root is selected by a combination of the MSTI priority value and  
the bridge’s MAC address. The MSTI priority is analogous to the RSTP  
bridge priority value. Where they differ is that while the RSTP bridge  
priority is used to determine the root bridge for an entire bridged network,  
MSTI priority is used only to determine the regional root for a particular  
MSTI.  
The range for this parameter is the same as the RSTP bridge priority; from  
0 to 61,440 in sixteen increments of 4,096. To set the parameter, you  
specify the increment that represents the desired MSTI priority value.  
Table 19 on page 219 lists the increments.  
Region Following are several points to remember about regions.  
Guidelines  
ˆ A network can contain any number of regions and a region can contain  
any number of AT-9400 Switches.  
ˆ The AT-9400 Switch can belong to only one region at a time.  
ˆ A region can contain any number of VLANs.  
ˆ All of the bridges in a region must have the same configuration name,  
revision level, VLANs, and VLAN to MSTI associations.  
ˆ An MSTI cannot span multiple regions.  
ˆ Each MSTI must have a regional root for locating loops in the instance.  
MSTIs can share the same regional root or have different roots. A  
regional root is determined by the MSTI priority value and a bridge’s  
MAC address.  
ˆ The regional root of a MSTI must be in the same region as the MSTI.  
Section V: Spanning Tree Protocols  
237  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 21: Multiple Spanning Tree Protocol  
Common and MSTP has a default spanning tree instance called the Common and  
Internal Spanning Tree (CIST). This instance has an MSTI ID of 0.  
Internal  
Spanning Tree  
(CIST)  
This instance has unique features and functions that make it different from  
the MSTIs that you create yourself. Firstly, you cannot delete this instance  
and you cannot change its MSTI ID.  
Secondly, when you create a new port-based or tagged VLAN, it is by  
default associated with the CIST and is automatically given an MSTI ID of  
0. The Default_VLAN is also associated by default with CIST.  
Another critical difference is that when you assign a VLAN to another  
MSTI, it still partially remains a member of CIST. This is because CIST is  
used by MSTP to communicate with other MSTP regions and with any  
RSTP and STP single-instance spanning trees in the network. MSTP uses  
CIST to participate in the creation of a spanning tree between different  
regions and between regions and single-instance spanning tree, to form  
one spanning tree for the entire bridged network.  
MSTP uses CIST to form the spanning tree of an entire bridged network  
because CIST can cross regional boundaries, while a MSTI cannot. If a  
port is a boundary port, that is, if it is connected to another region, that port  
automatically belongs solely to CIST, even if it was assigned to an MSTI,  
because only CIST is active outside of a region.  
As mentioned earlier, every MSTI must have a root bridge, referred to as a  
regional root, in order to locate loops that might exist within the instance.  
CIST must also have a regional root. However, the CIST regional root  
communicates with the other MSTP regions and single-instance spanning  
trees in the bridged network.  
The CIST regional root is set with the CIST Priority parameter. This  
parameter, which functions similar to the RSTP bridge priority value,  
selects the root bridge for the entire bridged network. If the AT-9400  
Switch has the lowest CIST Priority value among all the spanning tree  
bridges, it functions as the root bridge for all the MSTP regions and STP  
and RSTP single-instance spanning trees in the network.  
MSTP with STP MSTP is fully compatible with STP and RSTP. If a port on the AT-9400  
Switch running MSTP receives STP BPDUs, the port sends only STP  
BPDU packets. If a port receives RSTP BPDUs, the port sends MSTP  
BPDUs because RSTP can process MSTP BPDUs.  
and RSTP  
A port connected to a bridge running STP or RSTP is considered to be a  
boundary port of the MSTP region and the bridge as belonging to a  
different region.  
An MSTP region can be considered as a virtual bridge. The implication is  
that other MSTP regions and STP and RSTP single-instance spanning  
trees cannot discern the topology or constitution of a MSTP region. The  
only bridge they are aware of is the regional root of the CIST instance.  
238  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
         
AT-S63 Management Software Features Guide  
Summary of Guidelines  
Careful planning is essential for the successful implementation of MSTP.  
This section reviews all the rules and guidelines mentioned in earlier  
sections, and contains a few new ones:  
ˆ The AT-9400 Switch can support up to 16 spanning tree instances,  
including the CIST, at a time.  
ˆ A MSTI can contain any number of VLANs.  
ˆ A VLAN can belong to only one MSTI at a time.  
ˆ An MSTI ID can be from 1 to 15.  
ˆ The CIST ID is 0. You cannot change this value.  
ˆ A switch port can belong to more than one spanning tree instance at a  
time. This allows you to assign a port as an untagged and tagged  
member of VLANs that belong to different MSTIs. What makes this  
possible is a port’s ability to be in different MSTP states for different  
MSTIs simultaneously. For example, a port can be in the MSTP  
blocking state for one MSTI and the forwarding state for another  
spanning tree instance.  
ˆ A router or Layer 3 network device is required to forward traffic  
between VLANs.  
ˆ A network can contain any number of regions and a region can contain  
any number of AT-9400 Switches.  
ˆ The AT-9400 Switch can belong to only one region at a time.  
ˆ A region can contain any number of VLANs.  
ˆ All of the bridges in a region must have the same configuration name,  
revision level, VLANs, and VLAN to MSTI associations.  
ˆ An MSTI cannot span multiple regions.  
ˆ Each MSTI must have a regional root for locating loops in the instance.  
MSTIs can share the same regional root or have different roots. A  
regional root is determined by the MSTI priority value and a bridge’s  
MAC address.  
ˆ The regional root of a MSTI must be in the same region as the MSTI.  
ˆ The CIST must have a regional root for communicating with other  
regions and single-instance spanning trees.  
ˆ MSTP is compatible with STP and RSTP.  
ˆ A port transmits CIST information even when it is associated with  
another MSTI ID. However, in determining network loops, MSTI takes  
precedence over CIST. (This is explained more in “Associating VLANs  
Section V: Spanning Tree Protocols  
239  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 21: Multiple Spanning Tree Protocol  
Note  
The AT-S63 MSTP implementation complies fully with the new IEEE  
802.1s standard. Any other vendor’s fully compliant 802.1s  
implementation is interoperable with the AT-S63 implementation.  
240  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Associating VLANs to MSTIs  
Allied Telesis recommends that you assign all VLANs on a switch to an  
MSTI. You should not leave a VLAN assigned to just the CIST, including  
the Default_VLAN. This is to prevent the blocking of a port that should be  
in the forwarding state. The reason for this guideline is explained below.  
An MSTP BPDU contains the instance to which the port transmitting the  
packet belongs. By default, all ports belong to the CIST instance. So CIST  
is included in the BPDU. If the port is a member of a VLAN that has been  
assigned to another MSTI, that information is also included in the BPDU.  
This is illustrated in Figure 29. Port 8 in switch A is a member of a VLAN  
assigned to MSTI ID 7 while port 1 is a member of a VLAN assigned to  
MSTI ID 10. The BPDUs transmitted by port 8 to switch B would indicate  
that the port is a member of both CIST and MSTI 7, while the BPDUs from  
port 1 would indicate the port is a member of the CIST and MSTI 10.  
BPDU Packet  
Instances: CIST 0 and MSTI 10  
Port 1  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT  
-9424T/SP Gigabit Ethernet Switch  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT  
-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
LASER PRODUCT  
H
D
X
/
C
O
L
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
1
3
5
6
7
8
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
F
AUL  
T
L/A  
D/C  
L/A  
D/C  
F
AUL  
T
SFP  
1000 LINK  
L/A  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
SFP  
23  
SFP  
24  
/
ACT  
/
ACT  
MASTER  
RPS  
MASTER  
RPS  
Switch A  
23  
24  
23  
24  
Switch B  
POWER  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Port 8  
BPDU Packet  
Instances: CIST 0 and MSTI 7  
Figure 29. CIST and VLAN Guideline - Example 1  
At first glance, it might appear that because both ports belong to CIST, a  
loop would exist between the switches and that MSTP would block a port  
to stop the loop. However, within a region, MSTI takes precedence over  
CIST. When switch B receives a packet from switch A, it uses MSTI, not  
CIST, to determine whether a loop exists. And because both ports on  
switch A belong to different MSTIs, switch B determines that no loop  
exists.  
A problem can arise if you assign some VLANs to MSTIs while leaving  
others just to CIST. The problem is illustrated in Figure 30. The network is  
the same as the previous example. The only difference is that the VLAN  
containing port 8 on Switch A has not been assigned to an MSTI, and  
belongs only to CIST with its MSTI ID 0.  
Section V: Spanning Tree Protocols  
241  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 21: Multiple Spanning Tree Protocol  
BPDU Packet  
Instances: CIST 0 and MSTI 10  
Port 1  
Port 15  
19 21 23R  
CLASS  
LASER PRODUCT  
1
3
5
7
9
11  
13  
15  
17  
AT  
-9424T/SP Gigabit Ethernet Switch  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT  
-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
PORT ACTIVITY  
ACT 10/100 LINK  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
1
3
5
6
7
8
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
F
AUL  
T
L/A  
D/C  
L/A  
D/C  
F
AUL  
T
SFP  
1000 LINK  
L/A  
SFP  
1000 LINK  
L/A  
SFP  
SFP  
24  
SFP  
23  
SFP  
24  
/
ACT  
/
ACT  
MASTER  
RPS  
MASTER  
RPS  
Switch A  
Port 8  
23  
24  
23  
24  
Switch B  
POWER  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
10 12 14 16 18 20 22 24R  
2
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
23  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Port 4  
BPDU Packet  
Instances: CIST 0  
Figure 30. CIST and VLAN Guideline - Example 2  
When port 4 on switch B receives a BPDU, the switch notes the port  
sending the packet belongs only to CIST. Therefore, switch B uses CIST  
in determining whether a loop exists. The result would be that the switch  
detects a loop because the other port is also receiving BPDU packets from  
CIST 0. Switch B would block a port to cancel the loop.  
To avoid this issue, always assign all VLANs on a switch, including the  
Default_VLAN, to an MSTI. This guarantees that all ports on the switch  
have an MSTI ID and that helps to ensure that loop detection is based on  
MSTI, not CIST.  
242  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Connecting VLANs Across Different Regions  
Special consideration needs to be taken into account when you connect  
different MSTP regions or an MSTP region and a single-instance STP or  
RSTP region. Unless planned properly, VLAN fragmentation can occur  
between the VLANS of your network.  
As mentioned previously, only the CIST can span regions. A MSTI cannot.  
Consequently, you may run into a problem if you use more than one  
physical data link to connect together various parts of VLANs that reside in  
bridges in different regions. The result can be a physical loop, which  
spanning tree disables by blocking ports.  
This is illustrated in Figure 31. The example show two switches, each  
residing in a different region. Port 1 in switch A is a boundary port. It is an  
untagged member of the Accounting VLAN, which has been associated  
with MSTI 4. Port 16 is a tagged and untagged member of three different  
VLANs, all associated to MSTI 12.  
If both switches were a part of the same region, there would be no problem  
because the ports reside in different spanning tree instances. However,  
the switches are part of different regions and MSTIs do not cross regions.  
Consequently, the result is that spanning tree would determine that a loop  
exists between the regions, and Switch B would block a port.  
Port 5  
MSTI 4  
Region 1  
Region 2  
VLAN (untagged) port: Accounting  
1
3
5
7
9
11  
13  
15  
17  
1
21  
23R  
AT  
-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT  
-9424T/GB Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
CLASS 1  
LASER PRODUCT  
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
F
AUL  
T
SFP  
1000 LINK  
L/A  
GBIC  
GBIC  
L/A  
D/C  
L/A  
D/C  
F
AUL  
T
SFP  
23  
SFP  
24  
GBIC  
1000 LINK  
L/A  
/
ACT  
MASTER  
RPS  
/
ACT  
MASTER  
RPS  
23  
24  
23  
24  
POWER  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
18  
20  
22  
24R  
23  
24  
Switch B  
Switch A  
Port 16  
MSTI 12  
VLAN (untagged port): Sales  
VLAN (tagged port): Presales  
VLAN (tagged port): Marketing  
Figure 31. Spanning Regions - Example 1  
There are several ways to address this issue. One is to have only one  
MSTP region for each subnet in your network.  
Section V: Spanning Tree Protocols  
243  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 21: Multiple Spanning Tree Protocol  
Another approach is to group those VLANs that need to span regions into  
the same MSTI. Those VLANs that do not span regions can be assigned  
to other MSTIs.  
Here is an example. Assume that you have two regions that contain the  
following VLANS:  
Region 1 VLANs  
Sales  
Presales  
Marketing  
Advertising  
Technical Support  
Product Management  
Project Management  
Accounting  
Region 2 VLANs  
Hardware Engineering  
Software Engineering  
Technical Support  
Product Management  
CAD Development  
Accounting  
The two regions share three VLANs: Technical Support, Product  
Management, and Accounting. You could group those VLANs into the  
same MSTI in each region. For instance, for Region 1 you might group the  
three VLANs in MSTI 11 and in Region 2 you could group them into MSTI  
6. After they are grouped, you can connect the VLANs across the regions  
using a link of tagged ports.  
244  
Section V: Spanning Tree Protocols  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Section VI  
Virtual LANs  
The chapters in this section discuss the various types of virtual LANs  
supported by the AT-9400 Switch. The chapters include:  
Section VI: Virtual LANs  
245  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
246  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 22  
Port-based and Tagged VLANs  
This chapter contains overview information about port-based and tagged  
virtual LANs (VLANs). This chapter contains the following sections:  
Section VI: Virtual LANs  
247  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 22: Port-based and Tagged VLANs  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
248  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
A VLAN is a group of ports on an Ethernet switch that form a logical  
Ethernet segment. The ports of a VLAN form an independent traffic  
domain where the traffic generated by the nodes of a VLAN remains within  
the VLAN.  
With VLANs, you can segment your network through the switch’s AT-S63  
Management Software and so be able to group nodes with related  
functions into their own separate, logical LAN segments. These VLAN  
groupings can be based on similar data needs or security requirements.  
For example, you could create separate VLANs for the different  
departments in your company, such as one for Sales and another for  
Accounting.  
VLANs offer several important benefits:  
ˆ Improved network performance  
Network performance often suffers as networks grow in size and as  
traffic increases. The more nodes on each LAN segment vying for  
bandwidth, the greater the likelihood overall network performance  
decreases.  
VLANs improve network perform because VLAN traffic stays within the  
VLAN. The nodes of a VLAN receive traffic only from nodes of the  
same VLAN. This reduces the need for nodes to handle traffic not  
destined for them. It also frees up bandwidth within all the logical  
workgroups.  
In addition, because each VLAN constitutes a separate broadcast  
domain, broadcast traffic remains within the VLAN. This too can  
improve overall network performance.  
ˆ Increased security  
Because network traffic generated by a node in a VLAN is restricted  
only to the other nodes of the same VLAN, you can use VLANs to  
control the flow of packets in your network and prevent packets from  
flowing to unauthorized end nodes.  
ˆ Simplified network management  
VLANs can also simplify network management. Before the advent of  
VLANs, physical changes to the network often had to been made at  
the switches in the wiring closets. For example, if an employee  
changed departments, changing the employee’s LAN segment  
assignment often required a change to the wiring at the switch.  
But with VLANS, you can change the LAN segment assignment of an  
end node connected to the switch using the switch’s AT-S63  
Section VI: Virtual LANs  
249  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 22: Port-based and Tagged VLANs  
Management Software. You can change the VLAN memberships  
through the management software without moving the workstations  
physically, or changing group memberships by moving cables from  
one switch port to another.  
In addition, a virtual LAN can span more than one switch. This means  
that the end nodes of a VLAN do not need to be connected to the  
same switch and so are not restricted to being in the same physical  
location.  
The AT-9400 Switch supports the following types of VLANs you can  
create yourself:  
ˆ Port-based VLANs  
ˆ Tagged VLANs  
These VLANs are described in the following sections.  
250  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Port-based VLAN Overview  
As explained in “Overview” on page 249, a VLAN consists of a group of  
ports on one or more Ethernet switches that form an independent traffic  
domain. Traffic generated by the end nodes of a VLAN remains within the  
VLAN and does not cross over to the end nodes of other VLANs unless  
there is an interconnection device, such as a router or Layer 3 switch.  
A port-based VLAN is a group of ports on a Gigabit Ethernet Switch that  
form a logical Ethernet segment. Each port of a port-based VLAN can  
belong to only one VLAN at a time.  
A port-based VLAN can have as many or as few ports as needed. The  
VLAN can consist of all the ports on an Ethernet switch, or just a few ports.  
A port-based VLAN also can span switches and consist of ports from  
multiple Ethernet switches.  
Note  
The AT-9400 Switch is preconfigured with one port-based VLAN. All  
ports on the switch are members of this VLAN, called the  
Default_VLAN.  
The parts that make up a port-based VLAN are:  
ˆ VLAN name  
ˆ VLAN Identifier  
ˆ Untagged ports  
ˆ Port VLAN Identifier  
VLAN Name To create a port-based VLAN, you must give it a name. The name should  
reflect the function of the network devices that are be members of the  
VLAN. Examples include Sales, Production, and Engineering.  
VLAN Identifier Every VLAN in a network must have a unique number assigned to it. This  
number is called the VLAN identifier (VID). This number uniquely identifies  
a VLAN in the switch and the network.  
If a VLAN consists only of ports located on one physical switch in your  
network, you assign it a VID different from all other VLANs in your network.  
If a VLAN spans multiple switches, then the VID for the VLAN on the  
different switches should be the same. The switches are then able to  
recognize and forward frames belonging to the same VLAN even though  
the VLAN spans multiple switches.  
For example, if you had a port-based VLAN titled Marketing that spanned  
Section VI: Virtual LANs  
251  
Download from Www.Somanuals.com. All Manuals Search And Download.  
           
Chapter 22: Port-based and Tagged VLANs  
three AT-9400 Switches, you would assign the Marketing VLAN on each  
switch the same VID.  
You can assign this number manually or allow the AT-S63 Management  
Software to do it automatically. If you allow the management software to  
do it automatically, it selects the next available VID. This is acceptable  
when you are creating a new, unique VLAN.  
If you are creating a VLAN on a switch that will be part of a larger VLAN  
that spans several switch, then you will need to assign the number  
yourself so that the VLAN has the same VID on all switches.  
Untagged Ports You need to specify which ports on the switch are to be members of a  
port-based VLAN. Ports in a port-based VLAN are referred to as untagged  
ports and the frames received on the ports as untagged frames. The  
names derive from the fact that the frames received on a port will not  
contain any information that indicates VLAN membership, and that VLAN  
membership will be determined solely by the port’s PVID. (There is  
another type of VLAN where VLAN membership is determined by  
information within the frames themselves, rather than by a port’s PVID.  
This type of VLAN is explained in “Tagged VLAN Overview” on page 257.)  
A port on a switch can be an untagged member of only one port-based  
VLAN at a time. An untagged port cannot be assigned to two port-based  
VLANs simultaneously.  
Port VLAN Each port in a port-based VLAN must have a port VLAN identifier (PVID).  
The switch associates a frame to a port-based VLAN by the PVID  
Identifier  
assigned to the port on which the frame is received, and forwards the  
frame only to those ports with the same PVID. Consequently, all ports of a  
port-based VLAN must have the same PVID. Additionally, the PVID of the  
ports in a VLAN must match the VLAN’s VID.  
For example, if you were creating a port-based VLAN on a switch and you  
had assigned the VLAN the VID 5, the PVID for each port in the VLAN  
would need to be assigned the value 5.  
Some switches and switch management programs require that you assign  
the PVID value for each port manually. However, the AT-S63  
Management Software performs this task automatically. The software  
automatically assigns a PVID to a port, making it identical to the VID of the  
VLAN to which the port is a member, when you assign the port as an  
untagged member to a VLAN.  
252  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
AT-S63 Management Software Features Guide  
Guidelines to Below are the guidelines to creating a port-based VLAN.  
Creating a Port-  
based VLAN  
ˆ Each port-based VLAN must be assigned a unique VID. If a particular  
VLAN spans multiples switches, each part of the VLAN on the different  
switches should be assigned the same VID.  
ˆ A port can be an untagged member of only one port-based VLAN at a  
time.  
ˆ The PVID of a port is identical to the VID of the VLAN where the port is  
an untagged member. The PVID value is automatically assigned by  
the AT-S63 Management Software.  
ˆ A port-based VLAN that spans multiple switches requires a port on  
each switch where the VLAN is located to function as an  
interconnection between the switches where the various parts of the  
VLAN reside.  
ˆ The switch can support up to a total of 4094 port-based, tagged,  
protected ports, and MAC address-based VLANs.  
ˆ A port set to the 802.1x authenticator or supplicant role must be  
changed to the 802.1x none role before you can change its untagged  
VLAN assignment. After the VLAN assignment is made, the port’s role  
can be changed back again to authenticator or supplicant, if desired.  
ˆ You cannot delete the Default VLAN from the switch.  
ˆ Deleting an untagged port from the Default VLAN without assigning it  
to another VLAN results in the port being an untagged member of no  
VLAN.  
Drawbacks of There are several drawbacks to port-based VLANs:  
Port-based  
ˆ It is not easy to share network resources, such as servers and printers,  
VLANs  
across multiple VLANs. A router or Layer 3 switch must be added to  
the network to provide a means for interconnecting the port-based  
VLANs. The introduction of a router into your network could create  
security issues from unauthorized access to your network.  
ˆ A VLAN that spans several switches requires a port on each switch for  
the interconnection of the various parts of the VLAN. For example, a  
VLAN that spans three switches would require one port on each switch  
to interconnect the various sections of the VLAN. In network  
configurations where there are many individual VLANs that span  
switches, many ports could end up being used ineffectively just to  
interconnect the various VLANs.  
Section VI: Virtual LANs  
253  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 22: Port-based and Tagged VLANs  
Port-based Figure 32 illustrates an example of one AT-9424T/SP Gigabit Ethernet  
Switch with three port-based VLANs. (For purposes of the following  
examples, the Default_VLAN is not shown.)  
Example 1  
Engineering VLAN  
(VID 3)  
Production VLAN  
(VID 4)  
Sales VLAN  
(VID 2)  
1
2
3
4
5
6
7
8
9
11 13 15 17 19 21 23  
AT-9424T/SP Gigabit Ethernet Switch  
10 12 14 16 18 20 22 24  
WAN  
Router  
Figure 32. Port-based VLAN - Example 1  
The table below lists the port assignments for the Sales, Engineering, and  
Production VLANs on the switch.  
Sales VLAN (VID 2)  
Engineering VLAN  
(VID 3)  
Production VLAN  
(VID 4)  
AT-9424T/SP  
Switch  
Ports 1, 3 - 5  
(PVID 2)  
Ports 9, 11 - 13  
(PVID 3)  
Ports 17 - 19, 21  
(PVID 4)  
Each VLAN has a unique VID. This number is assigned when you create a  
VLAN.  
The ports have been assigned PVID values. A port’s PVID is assigned  
automatically by the AT-S63 Management Software when you create the  
VLAN. The PVID of a port is the same as the VID to which the port is an  
untagged member.  
254  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
In the example, each VLAN has one port connected to the router. The  
router interconnects the various VLANs and functions as a gateway to the  
WAN.  
Port-based Figure 33 illustrates more port-based VLANs. In this example, two VLANs,  
Sales and Engineering, span two AT-9400 Switches Gigabit Ethernet  
switches.  
Example 2  
Engineering VLAN  
(VID 3)  
Sales VLAN  
(VID 2)  
Production VLAN  
(VID 4)  
1
2
3
4
5
6
7
8
9
11 13 15 17 19 21 23  
AT-9424T/SP Gigabit Ethernet Switch  
10 12 14 16 18 20 22 24  
WAN  
Router  
1
2
3
4
5
6
7
8
9
11 13 15 17 19 21 23  
AT-9424T/GB Gigabit Ethernet Switch  
10 12 14 16 18 20 22 24  
Sales VLAN  
(VID 2)  
Engineering VLAN  
(VID 3)  
Figure 33. Port-based VLAN - Example 2  
Section VI: Virtual LANs  
255  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 22: Port-based and Tagged VLANs  
The table below lists the port assignments for the Sales, Engineering, and  
Production VLANs on the switches:  
Sales VLAN  
(VID 2)  
Engineering VLAN  
(VID 3)  
Production VLAN  
(VID 4)  
AT-9424T/SP  
Switch (top)  
Ports 1 - 6  
(PVID 2)  
Ports 9 - 13  
(PVID 3)  
Ports 17, 19 - 21  
(PVID 4)  
AT-9424T/GB  
Switch (bottom)  
Ports 2 - 4, 6, 8  
(PVID 2)  
Ports 16, 18-20, 22  
(PVID 3)  
none  
ˆ Sales VLAN - This VLAN spans both switches. It has a VID value of 2  
and consists of six untagged ports on the top switch and five untagged  
ports on the bottom switch.  
The two parts of the VLAN are connected by a direct link from port 4  
on the top switch to port 3 on the bottom switch. This direct link allows  
the two parts of the Sales VLAN to function as one logical LAN  
segment.  
Port 6 on the top switch connects to the router. This port allows the  
Sales VLAN to exchange Ethernet frames with the other VLANs and to  
access the WAN.  
ˆ Engineering VLAN - The workstations of this VLAN are connected to  
ports 9 to 13 on the top switch and ports 16, 18 to 20, and 22 on the  
bottom switch.  
Because this VLAN spans multiple switches, it needs a direct  
connection between its various parts to provide a communications  
path. This is provided in the example with a direct connection from port  
10 on the top switch to port 19 on the bottom switch.  
This VLAN uses port 12 on the top switch as a connection to the router  
and the WAN.  
ˆ Production VLAN - This is the final VLAN in the example. It has the  
VLAN of 4 and its ports have been assigned the PVID also of 4.  
The nodes of this VLAN are connected only to the top switch. So this  
VLAN does not require a direct connection to the bottom switch.  
However, it uses port 20 as a connection to the router.  
256  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Tagged VLAN Overview  
The second type of VLAN supported by the AT-S63 Management  
Software is the tagged VLAN. VLAN membership in a tagged VLAN is  
determined by information within the frames that are received on a port.  
This differs from a port-based VLAN, where the PVIDs assigned to the  
ports determine VLAN membership.  
The VLAN information within an Ethernet frame is referred to as a tag or  
tagged header. A tag, which follows the source and destination addresses  
in a frame, contains the VID of the VLAN to which the frame belongs (IEEE  
802.3ac standard). As explained earlier in this chapter in “VLAN Identifier”  
on page 251, this number uniquely identifies each VLAN in a network.  
When a switch receives a frame with a VLAN tag, referred to as a tagged  
frame, the switch forwards the frame only to those ports that share the  
same VID.  
A port to receive or transmit tagged frames is referred to as a tagged port.  
Any network device connected to a tagged port must be IEEE 802.1Q-  
compliant. This is the standard that outlines the requirements and  
standards for tagging. The device must be able to process the tagged  
information on received frames and add tagged information to transmitted  
frames.  
The benefit of a tagged VLAN is that the tagged ports can belong to more  
than one VLAN at one time. This can greatly simplify the task of adding  
shared devices to the network. For example, a server can be configured to  
accept and return packets from many different VLANs simultaneously.  
Tagged VLANs are also useful where multiple VLANs span across  
switches. You can use one port per switch to connect all VLANs on the  
switch to another switch.  
The IEEE 802.1Q standard describes how this tagging information is used  
to forward the traffic throughout the switch. The handling of frames tagged  
with VIDs coming into a port is straightforward. If the incoming frame’s VID  
tag matches one of the VIDs of a VLAN of which the port is a tagged  
member, the frame is accepted and forwarded to the appropriate ports. If  
the frame’s VID does not match any of the VLANs that the port is a  
member of, the frame is discarded.  
The parts of a tagged VLAN are much the same as those for a port-based  
VLAN. They are:  
ˆ VLAN Name  
ˆ VLAN Identifier  
ˆ Tagged and Untagged Ports  
Section VI: Virtual LANs  
257  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 22: Port-based and Tagged VLANs  
ˆ Port VLAN Identifier  
Note  
For explanations of VLAN name and VLAN identifier, refer back to  
Tagged and You need to specify which ports will be members of the VLAN. In the case  
of a tagged VLAN, it is usually a combination of both untagged ports and  
tagged ports. You specify which ports are tagged and which untagged  
when you create the VLAN.  
Untagged Ports  
An untagged port, whether a member of a port-based VLAN or a tagged  
VLAN, can be in only one VLAN at a time. However, a tagged port can be  
a member of more than one VLAN. A port can also be an untagged  
member of one VLAN and a tagged member of different VLANs  
simultaneously.  
Port VLAN As explained earlier in the discussion on port-based VLANs, the PVID of a  
port determines the VLAN where the port is an untagged member.  
Identifier  
Because a tagged port determines VLAN membership by examining the  
tagged header within the frames that it receives and not the PVID, you  
could conclude that there is no need for a PVID. However, the PVID is  
used if a tagged port receives an untagged frame—a frame without any  
tagged information. The port forwards the frame based on the port’s PVID.  
This is only in cases where an untagged frame arrives on a tagged port.  
Otherwise, the PVID on a tagged port is ignored.  
Guidelines to Below are the guidelines to creating a tagged VLAN.  
Creating a  
Tagged VLAN  
ˆ Each tagged VLAN must be assigned a unique VID. If a particular  
VLAN spans multiple switches, each part of the VLAN on the different  
switches must be assigned the same VID.  
ˆ A tagged port can be a member of multiple VLANs.  
ˆ An untagged port can be an untagged member of only one VLAN at a  
time.  
ˆ The switch can support up to a total of 4094 port-based, tagged,  
protected ports, and MAC address-based VLANs.  
258  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
         
AT-S63 Management Software Features Guide  
Tagged VLAN Figure 34 illustrates how tagged ports can be used to interconnect IEEE  
802.1Q-based products.  
Example  
Engineering VLAN  
(VID 3)  
Sales VLAN  
(VID 2)  
Production VLAN  
(VID 4)  
Legacy Server  
1
2
3
4
5
6
7
8
9
11 13 15 17 19 21 23  
AT-9424T/SP Gigabit Ethernet Switch  
10 12 14 16 18 20 22 24  
IEEE 802.1Q-compliant  
Server  
WAN  
Router  
1
2
3
4
5
6
7
8
9
11 13 15 17 19 21 23  
AT-9424T/GB Gigabit Ethernet Switch  
10 12 14 16 18 20 22 24  
Engineering VLAN  
(VID 3)  
Sales VLAN  
(VID 2)  
Figure 34. Example of a Tagged VLAN  
Section VI: Virtual LANs  
259  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 22: Port-based and Tagged VLANs  
The port assignments for the VLANs are as follows:  
Engineering VLAN  
(VID 3)  
Production VLAN  
(VID 4)  
Sales VLAN (VID 2)  
Untagged  
Ports  
Tagged  
Ports  
Untagged  
Ports  
Tagged  
Ports  
Untagged  
Ports  
Tagged  
Ports  
AT-9424T/  
SP Switch  
(top)  
1, 3 to 5  
(PVID 2)  
2, 10  
9, 11 to 13  
(PVID 3)  
2, 10  
17, 19 to 21  
(PVID 4)  
2
AT-9424T/  
GB Switch  
(bottom)  
2, 4, 6, 8  
(PVID 2)  
9
16, 18, 20,  
22  
(PVID 3)  
9
none  
none  
This example is nearly identical to the “Port-based Example 2” on  
page 255. Tagged ports have been added to simplify network  
implementation and management.  
One of the tagged ports is port 2 on the top switch. This port has been  
made a tagged member of the three VLANs. It is connected to an IEEE  
802.1Q-compliant server, meaning the server can handle frames from  
multiple VLANs. Now all three VLANs can access the server without going  
through a router or other interconnection device.  
It is important to note that even though the server is accepting frames from  
and transmitting frames to more than one VLAN, data separation and  
security remain.  
Two other tagged ports are used to simplify network design in the  
example. They are port 10 on the top switch and port 9 on the lower  
switch. These ports have been made tagged members of the Sales and  
Engineering VLANs so that they can carry traffic from both VLANs,  
simultaneously. These ports provide a common connection that enables  
different parts of the same VLAN to communicate with each other while  
maintaining data separation between VLANs.  
In comparison, the Sales and Engineering VLANs in the “Port-based  
Example 2” on page 255 each had to have its own individual network link  
between the switches to connect the different parts of the VLANs. But with  
tagged ports, you can use one data link to carry data traffic from several  
VLANs, while still maintaining data separation and security. The tagged  
frames, when received by the switch, are delivered only to those ports that  
belong to the VLAN from which the tagged frame originated.  
260  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 23  
GARP VLAN Registration Protocol  
This chapter describes the GARP VLAN Registration Protocol (GVRP) and  
contains the following sections:  
Section VI: Virtual LANs  
261  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 23: GARP VLAN Registration Protocol  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
262  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Overview  
The GARP VLAN Registration Protocol (GVRP) allows network devices to  
share VLAN information. The main purpose of GVRP is to allow switches  
to automatically discover some of the VLAN information that would  
otherwise need to be manually configured in each switch. This is helpful in  
networks where VLANs span more than one switch. Without GVRP, you  
must manually configure your switches to ensure that the various parts of  
a VLAN can communicate across the different switches. GVRP, which is  
an application of the Generic Attribute Registration Protocol (GARP), does  
this for you automatically.  
The AT-S63 Management Software uses GVRP protocol data units  
(PDUs) to share VLAN information among GVRP-active devices. The  
PDUs contain the VID numbers of the VLANs on the switch. A PDU  
contains the VIDs of all the VLANs on the switch, not just the VID of which  
the transmitting port is a member.  
When a switch receives a GVRP PDU on a port, it examines the PDU to  
determine the VIDs of the VLANs on the device that sent it. It then does  
the following:  
ˆ If a VLAN does not exist on the switch, it creates the VLAN and adds  
the port as a tagged member to the VLAN. A VLAN created by GVRP  
is called a dynamic GVRP VLAN.  
ˆ If the VLAN already exists on the switch but the port is not a member of  
it, the switch adds the port as a tagged member. A port that has been  
added by GVRP to a static VLAN (that is a user-created VLAN) is  
called a dynamic GVRP port.  
You cannot modify a dynamic GVRP VLAN. After it is created, only GVRP  
can modify or delete it. A dynamic GVRP VLAN exists only so long as  
there are active nodes in the network that belong to the VLAN. If all nodes  
of a dynamic GVRP VLAN are shut down and there are no active links, the  
VLAN is deleted from the switch.  
A dynamic GVRP port in a static VLAN remains a member of the VLAN as  
long as there are active VLAN members. If all members of the VLAN  
become inactive or there are no active links, GVRP removes the dynamic  
port from the VLAN, but does not delete the VLAN if the VLAN is a static  
VLAN.  
Section VI: Virtual LANs  
263  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 23: GARP VLAN Registration Protocol  
Figure 35 provides an example of how GVRP works.  
Port 4  
Port 1  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT  
-9424T/GB Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT  
-9424T/GB Gigabit Ethernet Switch  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
PORT ACTIVITY  
ACT 10/100 LINK  
GBIC  
GBIC  
L/A  
D/C  
L/A  
D/C  
F
AUL  
T
CLASS 1  
LASER PRODUCT  
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
GBIC  
1000 LINK  
L/A  
/
ACT  
MASTER  
RPS  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
8
9
11 13 15 17 19 21 23R  
23  
24  
GBIC  
GBIC  
L/A  
D/C  
L/A  
D/C  
F
AUL  
T
GBIC  
1000 LINK  
L/A  
Port 3  
POWER  
/
ACT  
2
4
8
10 12 14 16 18 20 22 24R  
MASTER  
RPS  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
23  
24  
POWER  
2
4
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT  
-9424T/GB Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
GBIC  
GBIC  
L/A  
D/C  
L/A  
D/C  
F
AUL  
T
GBIC  
1000 LINK  
L/A  
/
ACT  
MASTER  
RPS  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
23  
24  
Port 2  
Switch #2  
Switch #1  
Static VLAN  
Sales VID=11  
Switch #3  
Static VLAN  
Sales VID=11  
Figure 35. GVRP Example  
Switches #1 and #3 contain the Sales VLAN, but switch #2 does not.  
Consequently, the end nodes of the two parts of the Sales VLANs are  
unable to communicate with each other.  
Without GVRP, you would need to configure switch #2 by creating the  
Sales VLAN on the switch and adding ports 2 and 3 as members of the  
VLAN. If you happen to have a large network with a large number of  
VLANs, such manual configurations can be cumbersome and time  
consuming.  
GVRP can make the configurations for you. Here is how GVRP would  
resolve the problem in the example.  
1. Port 1 on switch #1 sends a PDU to port 2 on switch #2, containing the  
VIDs of all the VLANs on the switch. One of the VIDs in the PDU would  
be that of the Sales VLAN, VID 11.  
2. Switch #2 examines the PDU it receives on port 2 and notes that it  
does not have a VLAN with a VID 11. So it creates the VLAN as a  
dynamic GVRP VLAN and assigns it a VID 11 and the name  
GVRP_VLAN_11. (The name of a dynamic GVRP VLAN has the prefix  
“GVRP_VLAN_”, followed by the VID number.) The switch then adds  
port 2, the port that received the PDU, as a tagged member of the  
VLAN.  
3. Switch #2 sends a PDU out port 3 containing all of the VIDs of the  
VLANs on the switch, including the new GVRP_VLAN_11 with its VID  
of 11. (Note that port 3 is not yet a member of the VLAN. Ports are  
added to VLANs when they receive, not send a PDU.)  
4. Switch #3 receives the PDU on port 4 and, after examining it, notes  
that one of the VLANs on switch #2 has the VID 11, which matches the  
VID of an already existing VLAN on the switch. So it does not create  
the VLAN because it already exists. It then determines whether the  
port that received the PDU, in this case port 4, is a member of the  
VLAN. If it is not a member, it automatically adds the port to the VLAN  
264  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
as an tagged dynamic GVRP port. If the port is already a member of  
the VLAN, then no change is made.  
5. Switch #3 sends a PDU out port 4 to switch #2.  
6. Switch #2 receives the PDU on port 3 and then adds the port as a  
tagged dynamic GVRP port to the dynamic GVRP_VLAN_11 VLAN.  
There is now a communications path for the end nodes of the Sales VLAN  
on switches #1 and #3. GVRP created the new GVRP_VLAN_11 dynamic  
GVRP VLAN with a VID of 11 on switch #2 and added ports 2 and 3 to the  
VLAN as tagged dynamic GVRP ports.  
Section VI: Virtual LANs  
265  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 23: GARP VLAN Registration Protocol  
Guidelines  
Following are guidelines to observe when using this feature:  
ˆ GVRP is supported with STP and RSTP, or without spanning tree.  
GVRP is not supported with MSTP.  
ˆ GVRP is supported when the switch is operating in the tagged VLAN  
mode, which is the VLAN mode for creating your own tagged and port-  
based VLANs.  
ˆ GVRP is not supported when the switch is operating in either of the  
multiple VLAN modes.  
ˆ Both ports that constitute a network link between the switch and the  
other device must be running GVRP.  
ˆ You cannot modify or delete a dynamic GVRP VLAN.  
ˆ You cannot remove a dynamic GVRP port from a static or dynamic  
VLAN.  
ˆ GVRP can only detect a VLAN where there are active nodes, or where  
at least one end node of a VLAN has established a valid link with a  
switch. GVRP will not be aware of a VLAN where there are no active  
end nodes or if no end nodes have established a link with the switch.  
ˆ Resetting a switch erases all dynamic GVRP VLANs and dynamic  
GVRP port assignments. The switch relearns the dynamic  
assignments as it receives PDUs from the other switches.  
ˆ GVRP has three timers that you can set: join timer, leave timer, and  
leave all timer. The values for these timers must be set the same on all  
switches running GVRP. Timers with different values on different  
switches can result in GVRP compatibility problems.  
ˆ You can convert dynamic GVRP VLANs and dynamic GVRP port  
assignments to static VLANs and static port assignments.  
ˆ The default port settings on the switch for GVRP is active, meaning  
that the ports participate in GVRP. Allied Telesis recommends  
disabling GVRP on those ports that are connected to GVRP-inactive  
devices, meaning devices that do not feature GVRP.  
ˆ PDUs are transmitted to only those switch ports where GVRP is  
enabled.  
266  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
GVRP and Network Security  
GVRP should be used with caution because it can expose your network to  
unauthorized access. A network intruder can access to restricted parts of  
the network by connecting to a switch port running GVRP and transmitting  
a bogus GVRP PDU containing VIDs of restricted VLANs. GVRP would  
make the switch port a member of the VLANs and that could give the  
intruder access to restricted areas of your network.  
To protect against this type of network intrusion, consider the following:  
ˆ Activating GVRP only on those switch ports that are connected to  
other devices that support GVRP. Do not activate GVRP on ports that  
are connected to GVRP-inactive devices.  
ˆ Converting all dynamic GVRP VLANs and dynamic GVRP ports to  
static assignments, and then turning off GVRP on all switches. This  
preserves the new VLAN assignments while protecting against  
network intrusion.  
Section VI: Virtual LANs  
267  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 23: GARP VLAN Registration Protocol  
GVRP-inactive Intermediate Switches  
If two GVRP-active devices are separated by a GVRP-inactive switch, the  
GVRP-active devices may not be able to share VLAN information. There  
are two issues involved.  
The first is whether the intermediate switch forwards the GVRP PDUs that  
it receives from the GVRP-active switches. GVRP PDUs are management  
frames, intended for a switch’s CPU. In all likelihood, a GVRP-inactive  
switch will discard the PDUs because it does not recognize them.  
The second issue is that even if the GVRP-inactive switch forwards GVRP  
PDUs, it will not create the VLANs, at least not automatically.  
Consequently, even if the GVRP-active switches receive the PDUs and  
create the necessary VLANs, the intermediate switch may block the VLAN  
traffic, unless you modify its VLANs and port assignments manually.  
268  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Generic Attribute Registration Protocol (GARP) Overview  
The following is a technical overview of GARP. An understanding of GARP  
may prove helpful when you use GVRP.  
The purpose of the Generic Attribute Registration Protocol (GARP) is to  
provide a generic framework whereby devices in a bridged LAN, for  
example end stations and switches, can register and deregister attribute  
values, such as VLAN Identifiers, with each other. In doing so, the  
attributes are propagated to devices in the bridged LAN, and these  
devices form a “reachability” tree that is a subset of an active topology. For  
a bridged LAN, the active topology is normally that created and maintained  
by the Spanning Tree Protocol (STP).  
To use GARP, a GARP application must be defined. The Layer 2 switch  
has one GARP application presently implemented, GVRP.  
The GARP application specifies what the attribute represents.  
GARP defines the architecture, rules of operation, state machines and  
variables for the registration and deregistration of attribute values. By  
itself, GARP is not directly used by devices in a bridged LAN. It is the  
applications of GARP that perform meaningful actions. The use of GVRP  
allows dynamic filter entries for VLAN membership to be distributed  
among the forwarding databases of VLAN-active switches.  
A GARP participant in a switch or an end station consists of a GARP  
application component, and a GARP Information Declaration (GID)  
component associated with each port of the switch. One such GARP  
participant exists per port, per GARP application. The GARP Information  
Propagation (GIP) component propagates information between GARP  
participants for the same application in a switch. Protocol exchanges take  
place between GARP participants by means of LLC Type 1 services, using  
the group MAC address and PDU format defined for the GARP application  
concerned.  
Every instance of a GARP application includes a database to store the  
values of the attributes. Within GARP, attributes are mapped to GID  
indexes.  
Section VI: Virtual LANs  
269  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 23: GARP VLAN Registration Protocol  
GARP architecture is shown in Figure 36.  
Switch  
GARP Participant  
GARP Participant  
GARP Application  
GID  
GARP Application  
GID  
GIP  
LLC  
LLC  
MAC Layer:  
Port 1  
MAC Layer:  
Port 2  
Figure 36. GARP Architecture  
The GARP application component of the GARP participant is responsible  
for defining the semantics associated with the parameter values and  
operators received in GARP PDUs, and for generating GARP PDUs for  
transmission. The application uses the GID component, and the state  
machines associated with the operation of GID, in order to control its  
protocol interactions.  
An instance of GID consists of the set of state machines that define the  
current registration and declaration state of all attribute values associated  
with the GARP participant. Separate state machines exist for the applicant  
and registrar. This is shown in Figure 37.  
270  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
GID  
Attribute ... state:  
Attribute C state:  
Attribute B state:  
Attribute A state:  
Applicant  
State  
Registrar  
State  
Figure 37. GID Architecture  
GARP registers and deregisters attribute values through GARP messages  
sent at the GID level. A GARP participant that wishes to make a  
declaration (an applicant registering an attribute value) sends a JoinIn or  
JoinEmpty message. An applicant that wishes to withdraw a declaration  
(deregistering an attribute value) sends a LeaveEmpty or LeaveIn  
message. Following the de-registration of an attribute value, the applicant  
sends a number of Empty messages. The purpose of the Empty message  
is to prompt other applicants to send JoinIn/JoinEmpty messages. For the  
GARP protocol to be resilient against multiple lost messages, a LeaveAll  
message is available. Timers are used in the state machines to generate  
events and control state transitions.  
The job of the applicant is twofold:  
ˆ To ensure that this participant’s declarations are registered by other  
participants’ registrars  
ˆ To ensure that other participants have a chance to redeclare (rejoin)  
after anyone withdraws a declaration (leaves).  
The applicant is therefore looking after the interests of all would-be  
participants. This allows the registrar to be very simple.  
The job of the registrar is to record whether an attribute is registered, in the  
process of being deregistered, or is not registered for an instance of GID.  
Section VI: Virtual LANs  
271  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 23: GARP VLAN Registration Protocol  
To control the applicant state machine, an applicant administrative control  
parameter is provided. This parameter determines whether or not the  
applicant state machine participates in GARP protocol exchanges. The  
default value has the applicant participating in the exchanges.  
To control the registrar state machine, a registrar administrative control  
parameter is provided. This parameter determines whether or not the  
registrar state machine listens to incoming GARP messages. The default  
value has the registrar listening to incoming GARP messages.  
The propagation of information between GARP participants for the same  
application in a switch is carried out by the GIP component. The operation  
of GIP is dependent upon STP being enabled on a port, because only  
ports in the STP Forwarding state are eligible for membership to the GIP  
connected ring. Ports in the GIP connected ring propagate GID Join and  
Leave requests to notify each other of attribute registrations and  
deregistrations. The operation of GIP allows ports in the switch to share  
information between themselves and the LANs/end stations to which the  
ports are connected.  
If a port enters the STP Forwarding state and the GARP application that  
the port belongs to is enabled, then the port is added to the GIP connected  
ring for the GARP application. All attributes registered by other ports in the  
GIP connected ring is propagated to the recently connected port. All  
attributes registered by the recently connected port is propagated to all  
other ports in the GIP connected ring.  
Similarly, if a port leaves the STP Forwarding state and the GARP  
application that the port belongs to is enabled, then the port is removed  
from the GIP connected ring for the GARP application. Prior to removal,  
GID leave requests are propagated to all other ports in the GIP connected  
ring if the port to be removed has previously registered an attribute and no  
other port in the GIP connected ring has registered that attribute. You can  
enable or disable GIP operations.  
272  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 24  
Multiple VLAN Modes  
This chapter describes the multiple VLAN modes. This chapter contains  
the following sections:  
Section VI: Virtual LANs  
273  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 24: Multiple VLAN Modes  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
274  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Overview  
The multiple VLAN modes are designed to simplify the task of configuring  
the switch in network environments that require a high degree of network  
segmentation. In a multiple VLAN mode, the ports on a switch are  
prohibited from forwarding traffic to each other and are only allowed to  
forward traffic to a user-designated uplink port. These configurations  
isolate the traffic on each port from all other ports, while providing access  
to the uplink port.  
The AT-S63 Management Software supports two types of multiple VLAN  
modes:  
ˆ 802.1Q-compliant Multiple VLAN mode  
ˆ Multiple VLAN mode (also referred to as non-802.1Q compliant  
Multiple VLAN mode)  
Each mode uses a different technique to isolate the ports and their traffic.  
The first method uses VLANs while the second uses port mapping. The  
uplink port is also different in each mode. In one the port is a tagged port  
and in the other untagged. This is explained in the following subsections.  
Note  
The multiple VLAN mode feature is supported only in single switch  
(i.e. edge switch) environments. This means that cascading of  
switches while in Multiple VLANs mode is not allowed.  
Configuring multiple VLANs on a cascaded switch can possibly  
result in disconnection of network paths between switches unless  
the port used to link the switch (being configured for multiple VLANs  
mode) is configured as uplink VLAN port.  
Configuring multiple VLANs on cascaded switches can also affect  
enhanced stacking because the master switch may not be able to  
detect member switches beyond the first cascaded switch.  
Section VI: Virtual LANs  
275  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 24: Multiple VLAN Modes  
802.1Q- Compliant Multiple VLAN Mode  
In this mode, each port is placed into a separate VLAN as an untagged  
port. The VLAN names and VID numbers are based on the port numbers.  
For example, the VLAN for port 4 is named Client_VLAN_4 and is given  
the VID of 4, the VLAN for port 5 is named Client_VLAN_5 and has a VID  
of 5, and so on.  
The VLAN configuration is accomplished automatically by the switch. After  
you select the mode and an uplink port, the switch forms the VLANs. It  
also assigns the PVID values as well. For example, the PVID for port 4 is  
assigned as 4, to match the VID of 4.  
A user-designated port on the switch functions as an uplink port, which  
can be connected to a shared device such as a router for access to a  
WAN. This port is placed as a tagged port in each VLAN. Thus, while the  
switch ports are separated from each other in their individual VLANs, they  
all have access to the uplink port.  
The uplink port also has its own VLAN, where it is an untagged member.  
This VLAN is called Uplink_VLAN.  
Note  
In 802.1Q Multiple VLAN mode, the device connected to the uplink  
port must be IEEE 802.1Q-compliant.  
An example of the 802.1Q-compliant VLAN mode is shown in Table 20.  
The table shows the VLANs on the AT-9400 Switch where port 22 has  
been selected as the uplink port.  
Table 20. 802.1Q-Compliant Multiple VLAN Example  
VLAN Name  
VID  
Untagged Port  
Tagged Port  
22  
Client_VLAN_1  
Client_VLAN_2  
Client_VLAN_3  
Client_VLAN_4  
Client_VLAN_5  
Client_VLAN_6  
Client_VLAN_7  
Client_VLAN_8  
Client_VLAN_9  
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9
22  
22  
22  
22  
22  
22  
22  
22  
276  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Table 20. 802.1Q-Compliant Multiple VLAN Example (Continued)  
VLAN Name  
VID  
10  
Untagged Port  
10  
Tagged Port  
22  
Client_VLAN_10  
Client_VLAN_11  
Client_VLAN_12  
Client_VLAN_13  
Client_VLAN_14  
Client_VLAN_15  
Client_VLAN_16  
Client_VLAN_17  
Client_VLAN_18  
Client_VLAN_19  
Client_VLAN_20  
Client_VLAN_21  
Uplink_VLAN  
11  
12  
13  
14  
15  
16  
17  
18  
19  
20  
21  
22  
23  
24  
11  
12  
13  
14  
15  
16  
17  
18  
19  
20  
21  
22  
23  
24  
22  
22  
22  
22  
22  
22  
22  
22  
22  
22  
22  
Client_VLAN_23  
Client_VLAN_24  
22  
22  
This highly segmented configuration is useful in situations where traffic  
generated by each end node or network segment connected to a port on  
the switch needs to be kept separate from all other network traffic, while  
still allowing access to an uplink to a WAN. Unicast traffic received by the  
uplink port is effectively directed to the appropriate port and end node and  
is not directed to any other port on the switch.  
The 802.1Q Multiple VLAN configuration is appropriate when the device  
connected to the uplink port is IEEE 802.1Q compatible, meaning that it  
can handle tagged packets.  
When you select the 802.1Q-compliant VLAN mode, you are asked to  
specify the uplink VLAN port. You can specify only one uplink port. The  
switch automatically configures the ports into the separate VLANs.  
Note  
The uplink VLAN is the management VLAN. Any remote  
management of the switch must be made through the uplink VLAN.  
Section VI: Virtual LANs  
277  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 24: Multiple VLAN Modes  
Non-802.1Q Compliant Multiple VLAN Mode  
Unlike the 802.1Q-compliant VLAN mode, which isolates port traffic by  
placing each port in a separate VLAN, this mode forms one VLAN with a  
VID of 1 that encompasses all ports. To establish traffic isolation, it uses  
port mapping. The result, however, is the same. Ports are permitted to  
forward traffic only to the designated uplink port and to no other port, even  
when they receive a broadcast packet.  
Another difference with this mode is that the uplink port is untagged.  
Consequently, you would use this mode when the device connected to the  
uplink port is not IEEE 802.1Q compatible, meaning that the device cannot  
handle tagged packets.  
Note  
When the uplink port receives a packet with a destination MAC  
address that is not in the MAC address table, the port broadcasts  
the packet to all switch ports. This can result in ports receiving  
packets that are not intended for them.  
Also note that a switch operating in this mode can be remotely  
managed through any port on the switch, not just the uplink port.  
278  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 25  
Protected Ports VLANs  
This chapter explains protected ports VLANs. It contains the following  
sections:  
Section VI: Virtual LANs  
279  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 25: Protected Ports VLANs  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from the following management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
280  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
The purpose of a protected ports VLAN is to allow multiple ports on the  
switch to share the same uplink port but not share traffic with each other.  
This feature has some of the same characteristics as the multiple VLAN  
modes described in the previous chapter, but it offers several advantages.  
One is that it provides more flexibility. With the multiple VLAN modes, you  
can select only one uplink port which is shared by all the other ports. Also,  
you are not allowed to modify the configuration. With protected ports  
VLANs, you can create LAN segments that consist of more than one port  
and you can specify multiple uplink ports.  
Another advantage is that the switch can support protected ports VLANs  
as well as port-based and tagged VLANs simultaneously, something that  
is not allowed with the multiple VLAN modes.  
An important concept of this feature is groups. A group is a selection of  
one or more ports that function as a LAN segment within the VLAN. The  
ports in each group are independent of the ports in the other groups of the  
VLAN. The ports of a group can share traffic only amongst themselves and  
with the uplink port, but not with ports in other groups of the VLAN.  
A protected ports VLAN can consist of two or more groups and a group  
can consist of one or more ports. The ports of a group can be either  
tagged or untagged.  
This type of VLAN also shares some common features with tagged  
VLANs, where one or more ports are shared by different LAN segments.  
But there are significant differences. First, all the ports in a tagged VLAN  
are considered a LAN segment, while the ports in a protected ports VLAN,  
though residing within a single VLAN, are subdivided into the smaller unit  
of groups, which represent the LAN segments.  
Second, a tagged VLAN, by its nature, contains one or more tagged ports.  
These are the ports that are shared among one or more tagged VLANs.  
The device connected to a tagged port must be 802.1Q compliant and it  
must be able to handle tagged packets.  
In contrast, the uplink port in a protected ports VLAN, which is shared by  
the ports in the different groups, can be either tagged or untagged. The  
device connected to it does not necessarily need to be 802.1Q compliant.  
Note  
For explanations of VIDs and tagged and untagged ports, refer to  
Section VI: Virtual LANs  
281  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 25: Protected Ports VLANs  
To create a protected ports VLAN, you perform many of the same steps  
that you do when you create a new port-based or tagged VLAN. You give  
it a name and a unique VID, and you indicate which of the ports will be  
tagged and untagged. What makes creating this type of VLAN different is  
that you must assign the ports of the VLAN to their respective groups.  
Following is an example of a protected ports VLAN. The first table lists the  
name of the VLAN, the VID, and the tagged and untagged ports. It also  
indicates which port will function as the uplink port, in this case port 22.  
The second table lists the different groups in the VLAN and the ports for  
each group.  
Name  
VID  
Internet_VLAN_1  
8
Untagged Ports in 1-10, 25  
VLAN  
Tagged Ports in  
VLAN  
none  
22  
Uplink Port(s)  
Group Number  
Port(s)  
1
2
3
4
5
6
1-2  
3
4
5-7  
8
9-10  
Allied Telesis recommends that you create tables similar to these before  
you create your own protected ports VLAN. Having the tables handy will  
make your job easier when the switch prompts you for this information.  
282  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Guidelines  
Following are the guidelines for implementing protected ports VLANS:  
ˆ A protected ports VLAN should contain a minimum of two groups. A  
protected ports VLAN of only one group can be replaced with a port-  
based or tagged VLAN instead.  
ˆ A protected ports VLAN can contain any number of groups.  
ˆ A group can contain any number of ports.  
ˆ The ports of a group can be tagged or untagged.  
ˆ Each group must be assigned a unique group number on the switch.  
The number can be from 1 to 256.  
ˆ A protected ports VLAN can contain more than one uplink port.  
ˆ The switch can support up to a total of 4094 port-based, tagged,  
protected ports, and MAC address-based VLANs.  
ˆ Uplink ports can be either tagged or untagged.  
ˆ Uplink ports can be shared among more than one protected ports  
VLAN, but only if they are tagged.  
ˆ A switch can contain a combination of port-based and tagged VLANs  
and protected ports VLANs.  
ˆ A port that is a member of a group in a protected ports VLAN cannot  
be a member of a port-based or tagged VLAN.  
ˆ A group can be a member of more than one protected ports VLAN at a  
time. However, the port members of the group must be identical in both  
VLANs and the ports must be tagged.  
ˆ You cannot create protected ports VLANs when the switch is operating  
in a multiple VLAN mode.  
ˆ An untagged port of a protected ports VLAN can not be made an  
untagged member of another protected ports VLAN until it is first  
removed from its current VLAN assignment and returned to the  
Default_VLAN.  
ˆ The switch must be set to the user-configured VLAN mode to support  
protected ports VLANs.  
Section VI: Virtual LANs  
283  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 25: Protected Ports VLANs  
284  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 26  
MAC Address-based VLANs  
This chapter contains overview information about MAC address-based  
VLANs. Sections in the chapter include:  
Section VI: Virtual LANs  
285  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 26: MAC Address-based VLANs  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
Not supported.  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from two of the management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
286  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
independent LAN segments within a network and are typically employed to  
improve network performance and security.  
The AT-S63 Management Software offers several different types of  
VLANs, including port-based, tagged, and protected ports. Membership in  
these VLANs is determined either by the port VLAN identifier (PVID)  
assigned to a port on a switch or, in the case of tagged traffic, by the VLAN  
identifier within the packets themselves.  
This chapter describes VLANs that are based on the source MAC  
addresses of the end nodes that are connected to the switch. With a MAC  
address-based VLAN, only those nodes whose source MAC addresses  
have been entered as members of the VLAN can share and access the  
VLAN resources. This is in contrast to a port-based or tagged VLAN where  
any node that has access a switch port can join a VLAN as a member.  
One of the principle advantages of this type of VLAN is that it can make it  
easier to manage network users that roam. These are users who access  
the network from different points at different times. The challenge for a  
network administrator is providing these users with the same resources  
regardless of the point at which they access the network. If you employed  
port-based or tagged VLANs for roaming users, you might have to  
reconfigure the VLANs, moving ports to and from different virtual LANs, so  
that the users always have access to the same network resources. But  
with a MAC address-based VLAN, the switch can assign a network user to  
the same VLAN and network resources regardless of the port from which  
the user accesses the network.  
Section VI: Virtual LANs  
287  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 26: MAC Address-based VLANs  
Egress Ports  
Implementing a MAC address-based VLAN involves more than entering  
the MAC addresses of the end nodes that are members of the VLAN. You  
must also designate the egress ports on the switch for the packets from  
the nodes. The egress ports define the limits of flooding of packets when a  
port receives a unicast packet with an unknown destination address (that  
is, an address that has not been learned by the MAC address table).  
Without knowing the egress ports, the switch would be forced to flood the  
packets on all switch ports, and that could result in a security violation  
where end nodes receive packets from other nodes that are in different  
VLANs.  
Table 21 illustrates a simple example of the mapping of addresses to  
egress ports for a MAC address-based VLAN of 6 nodes. The example  
consists of four workstations, a printer, and a server. For instance,  
Workstation 1 is connected to port 1 on the switch and is mapped to  
egress ports 5 for the server and 6 for the printer.  
Table 21. Mappings of MAC Addresses to Egress Ports Example  
Switch Egress  
MAC address  
End Node  
Port  
00:30:84:54:1A:45  
00:30:84:C3:5A:11  
00:30:84:22:67:17  
00:30:84:78:75:1C  
00:30:79:7A:11:10  
00:30:42:53:10:3A  
Workstation 1 (Port 1)  
Workstation 2 (Port 2)  
Workstation 3 (Port 3)  
Workstation 4 (Port 4)  
Server (Port 5)  
5, 6  
5, 6  
5, 6  
5, 6  
1-4  
Printer (Port 6)  
1-4  
Obviously, mapping source MAC addresses to egress ports can become  
cumbersome if you are dealing with a MAC address-based VLAN that  
encompasses a large number of ports and nodes. Fortunately, the egress  
ports of a VLAN are considered as a community and, as such, need only  
be designated as an egress port of one address in the VLAN to be  
considered an egress port of all the addresses.  
For instance, referring to the previous example, if workstation 1 sends a  
packet containing an unknown destination MAC address, the switch does  
not flood the packet to just ports 5 and 6, even though those are the  
designated egress ports for packets from workstation 1. Rather, it floods it  
out all egress ports assigned to all MAC address of the VLAN, except, of  
course, for the port where the packet was received. In the example the  
switch would flood the packet out ports 2 through 6.  
288  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
The community characteristic of egress ports relieves you from having to  
map each address to its corresponding egress port. You only need to be  
sure that all egress ports in a MAC address-based VLAN are represented  
at least once by being assigned to at least one address.  
It is also important to note that a MAC address must be assigned at least  
one egress port to be considered a member of a MAC address-based  
VLAN. VLAN membership of packets from a source MAC address not  
assigned any egress ports is determined by the PVID of the port where the  
packets are received.  
Because egress ports are considered as a community within a VLAN, you  
can simplify the mappings by assigning all of the egress ports to just one  
MAC address and, for the rest of the addresses, assigning just one port.  
This will make it easier to add or delete MAC addresses or egress ports  
from a VLAN. Here is how the example might look.  
Table 22. Revised Example of Mappings of MAC Addresses to Egress Ports  
MAC Address  
End Node  
Egress Port  
00:30:84:54:1A:45  
00:30:84:C3:5A:11  
00:30:84:22:67:17  
00:30:84:78:75:1C  
00:30:79:7A:11:10  
00:30:42:53:10:3A  
Workstation 1 (Port 1)  
Workstation 2 (Port 2)  
Workstation 3 (Port 3)  
Workstation 4 (Port 4)  
Server (Port 5)  
1-6  
1
1
1
1
Printer (Port 6)  
1
A switch can support more than one MAC-address VLAN at a time and a  
port can be an egress member of more than one VLAN. While this can  
prove useful in some situations, it can also result in VLAN leakage where  
the traffic of one VLAN crosses the boundary into other VLANs.  
The problem arises in the case of unknown unicast traffic. If the switch  
receives a packet from a member of a MAC address-based VLAN with an  
unknown destination address, it floods the packet on all egress ports of the  
VLAN. If the VLAN contains a port that is also serving as an egress port of  
another VLAN, the node connected to the port receives the flooded  
packets, even if it does not belong to the same VLAN as the node that  
generated the packet.  
Here’s an example. Assume that Port 4 on a switch has been designated  
an egress port of three MAC address-based VLANs. Any unknown unicast  
traffic that the switch receives that belong to any of the VLANs will be  
flooded out Port 4, even if there are no active members of that particular  
VLAN on the port. This means that whatever device is connected to the  
port receives the flooded traffic of all three VLANs.  
Section VI: Virtual LANs  
289  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 26: MAC Address-based VLANs  
If security is a major concern for your network, you might not want to  
assign a port as an egress port to more than one VLAN when planning  
your MAC address-based VLANs.  
When a packet whose source MAC address is part of a MAC address-  
based VLAN arrives on a port, the switch performs one of the following  
actions:  
ˆ If the packet’s destination MAC address is not in the MAC address  
table, the switch floods the packet out all egress ports of the VLAN,  
excluding the port where the packet was received.  
ˆ If the packet’s destination MAC address is in the MAC address table  
and if the port where the address was learned is one of the VLAN’s  
egress ports, the switch forwards the packet to the port.  
ˆ If the packet’s destination MAC address is in the MAC address table  
but the port where the address was learned is not one of the VLAN’s  
egress ports, the switch discards the packet.  
290  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
VLANs That Span Switches  
To create a MAC address-based VLAN that spans switches, you must  
replicate the MAC addresses of the VLAN nodes on all the switches where  
the VLAN exists. The same MAC address-based VLAN on different  
switches must have the same list of MAC addresses.  
Figure 38 illustrates an example of a MAC address-based VLAN that  
spans two AT-9400 Switches. The VLAN consists of three nodes on each  
switch. Table 23 on page 292 lists the details of the VLAN on the switches.  
Note that each VLAN contains the complete set of MAC addresses of all  
VLAN nodes along with the appropriate egress ports on the switches.  
MAC Addresses:  
Address_1  
Address_2  
Address_3  
1
2
3
4
5
6
7
8
9
11 13 15 17 19 21 23  
AT-9400 Switch (A)  
AT-9400 Switch (B)  
10 12 14 16 18 20 22 24  
1
2
3
4
5
6
7
8
9
11 13 15 17 19 21 23  
10 12 14 16 18 20 22 24  
MAC Addresses:  
Address_4  
Address_5  
Address_6  
Figure 38. Example of a MAC Address-based VLAN Spanning Switches  
Section VI: Virtual LANs  
291  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 26: MAC Address-based VLANs  
Table 23. Example of a MAC Address-based VLAN Spanning Switches  
Switch A  
Switch B  
VLAN Name: Sales  
MAC Address  
Address_1  
VLAN Name: Sales  
MAC Address  
Address_1  
Egress Ports  
Egress Ports  
1,3,4,5  
11,12,14,16  
Address_2  
1
1
1
1
1
Address_2  
11  
11  
11  
11  
11  
Address_3  
Address_3  
Address_4  
Address_4  
Address_5  
Address_5  
Address_6  
Address_6  
292  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
VLAN Hierarchy  
The switch’s management software employs a VLAN hierarchy when  
handling untagged packets that arrive on a port that is an egress port of a  
MAC address-based VLAN as well as an untagged port of a port-based  
VLAN. (A port can be a member of both types of VLANs at the same time.)  
The rule is that a MAC address-based VLAN takes precedence over that  
of a port-based VLAN.  
When an untagged packet arrives on a port, the switch first compares the  
source MAC address of the packet against the MAC addresses of all the  
MAC address-based VLANs on the device. If there is a match, the switch  
considers the packet as a member of the corresponding MAC address-  
based VLAN and not the port-based VLAN, and forwards it out the egress  
ports defined for the corresponding MAC address-based VLAN.  
If there is no match, the switch considers the packet as a member of the  
port-based VLAN and forwards the packet according to the PVID assigned  
to the port. For an explanation of a PVID, refer to “Port-based VLAN  
Section VI: Virtual LANs  
293  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 26: MAC Address-based VLANs  
Steps to Creating a MAC Address-based VLAN  
Here are the three main steps to creating a MAC address-based VLAN:  
1. Assign the VLAN a name and a VID. You must also set the VLAN type  
to MAC Based.  
2. Assign the MAC addresses to the VLAN.  
3. Add the egress ports to the MAC addresses.  
The steps must be performed in this order.  
294  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Guidelines  
Follow these guidelines when implementing a MAC address-based VLAN:  
ˆ MAC address-based VLANs are not supported on the AT-9408LC/SP,  
AT-9424T/GB and AT-9424T/SP switches.  
ˆ The switch can support up to a total of 4094 port-based, tagged,  
protected ports, and MAC address-based VLANs.  
ˆ The source nodes of this type of VLAN must send only untagged  
packets. A MAC address-based VLAN does not support tagged  
packets.  
ˆ The switch supports MAC address-based VLANs when operating in  
the user configured VLAN mode, which is the default setting for the  
switch. The switch must not be running a multiple VLAN mode.  
ˆ The egress ports of a VLAN are considered as a community. Assigning  
a port to one MAC address in a VLAN implicitly defines the port as an  
egress port of all the addresses in the same VLAN.  
ˆ A source MAC address must be assigned to at least one egress port to  
be considered part of a MAC address-based VLAN. Otherwise, VLAN  
membership is determined by the PVID of the port where the packets  
are received.  
ˆ A port can be an egress port of more than one MAC address-based  
VLAN.  
ˆ An egress port cannot be part of a port trunk.  
ˆ A MAC address can belong to only one MAC address-based VLAN at  
a time.  
ˆ A broadcast packet crosses VLAN boundaries when a port is an  
egress port of a MAC address-based VLAN and an untagged member  
of a port-based VLAN. Given that there is no way for the switch to  
determine the VLAN to which the broadcast packet belongs, it floods  
the packet on all ports of all affected VLANs.  
ˆ Entering a MAC address as part of a MAC address-based VLAN does  
not add the address to the MAC address table. The address appears in  
the MAC address table during the normal learning process of the  
switch.  
ˆ A MAC address-based VLAN is supported in an edge switch, where  
end nodes are connected directly to the switch, as well as in an  
intermediary switch, where the switch is connected to other Ethernet  
switches or hubs.  
ˆ The switch can support a total of 1024 MAC addresses in all its MAC  
address-based VLANs.  
ˆ A MAC address-based VLAN does not support multicast MAC  
addresses.  
Section VI: Virtual LANs  
295  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 26: MAC Address-based VLANs  
ˆ Egress ports cannot be part of a static or LACP trunk.  
ˆ Since this type of VLAN does not support tagged packets, it is not  
suitable in environments where a network device, such as a network  
server, needs to be shared between multiple VLANs.  
ˆ Ports 49 and 50 on the AT-9448Ts/XP switch cannot be designated as  
egress ports of a MAC address-based VLAN.  
ˆ SFP ports 45 to 48 on the AT-9448T/SP switch cannot be designated  
as egress ports of a MAC address-based VLAN.  
296  
Section VI: Virtual LANs  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Section VII  
Routing  
This section has the following chapters:  
Section VII: Internet Protocol Routing  
297  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
298  
Section VII: Internet Protocol Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 27: Internet Protocol Version 4 Packet Routing  
Supported Platforms  
This feature is supported on the following switches:  
ˆ Layer 2+ Models  
Not supported  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
Note  
You can create one routing interface on the Layer 2+ models and a  
stack of Basic Layer 3 switches to serve as the IP configuration for  
the device. For further information, refer to “Routing Interfaces and  
This feature can be managed from the following management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
Note  
You can configure routing interfaces from the menus interface, but  
not static routes or RIP. To configure all the feature’s components,  
you must use the command line interface.  
300  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
This section contains an overview of the IPv4 routing feature on the  
AT-9400 Switch. It begins with an explanation of the following available  
routing methods:  
ˆ Routing interfaces  
ˆ Static routes  
ˆ RIP version 1 and 2  
A routing interface is a logical connection to a local network or subnet for  
the purpose of routing IPv4 packets. Interfaces route packets between the  
local networks and subnets directly connected to the switch and are  
independent of static routes and RIP. In some limited network topologies  
where there are no remote networks or subnets, you may be able to meet  
the routing requirements of the IPv4 packets on your network with just  
In order for the switch to route packets to a remote destination (i.e., a  
network or subnet not directly connected to the switch), there must be a  
route to the destination in the routing table of the switch. A route consists  
of the IP address of the remote destination and the IP address of the next  
hop to reaching the destination.  
One method for specifying a route to a remote destination is to enter it  
manually. This type of route is referred to as a static route. A static route  
contains the IP addresses of the remote destination and the next hop. You  
can also create a static route for packets with an unknown destination  
network or subnet. This type of route is referred to as a default route. For  
background information on static routes and the default route, refer to  
A switch can automatically learn routes to remote destinations with the  
Routing Information Protocol (RIP). This protocol allows the routers of a  
network to automatically share their routes by broadcasting their routing  
tables to each other. The AT-9400 Switch supports versions 1 and 2 of this  
routing protocol. This feature is explained in “Routing Information Protocol  
This overview also contains an explanation of the role played by interfaces  
with some of the management features of the switch, and how those  
features are dependent on there being at least one interface on the switch.  
A few examples of the management functions include uploading and  
downloading files to the switch using a TFTP server and the enhanced  
stacking feature. For information, refer to “Routing Interfaces and  
Section VII: Routing  
301  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 27: Internet Protocol Version 4 Packet Routing  
At the end of this overview are two examples that illustrate the sequence  
of commands to implementing the features described in this chapter. You  
can refer there to see how the commands are used in practice. The  
In the following discussions, unless stated otherwise the term “remote  
destination” refers to a network or subnet that is not directly connected to  
the switch.  
302  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Routing Interfaces  
The IPv4 packet routing feature on the switch is built on the foundation of  
the routing interface. An interface functions as a logical connection to a  
subnet that allows the egress and ingress of IPv4 packets to the subnet  
from other local and remote networks, subnets, and nodes.  
Interfaces are an independent routing function. They are not dependent on  
static routes or RIP to pass IPv4 traffic among themselves on a switch. A  
switch automatically begins to route IPv4 packets among its local subnets  
as soon as two or more interfaces have been defined on the device.  
In order for a switch to route IPv4 traffic among it local subnets, it must  
have a routing interface on each subnet. You create an interface by  
assigning it a unique IP address of the subnet and indicating the VLAN  
where the subnet resides.  
Interfaces also function as anchor points for static routes. A static route  
defines the next hop to a remote destination. To create a static route to a  
remote destination, you add it to the interface on the switch where the next  
hop to the remote destination is located.  
Interfaces also act as anchor points for RIP. You can add RIP to the  
interfaces so that the switch automatically learns routes to remote  
destinations by sharing its routing information with the neighboring routers.  
In some limited network topologies, you might be able to meet the routing  
requirements of the IPv4 packets of your network with just routing  
interfaces. This would assume, of course, that the switch is directly  
connected to all of the networks or subnets of your network and that there  
are no remote destinations that would require static routes or RIP.  
Here are several other items to note concerning routing interfaces on the  
AT-9400 Switch:  
ˆ The switch can support up to 512 interfaces at one time, which means  
it can route the IPv4 traffic on up to 512 local subnets and networks.  
ˆ A single VLAN on a switch can contain up to sixteen interfaces.  
ˆ The AT-9408LC/SP, AT-9424T/GB, and AT-9424T/SP switches do not  
support the IPv4 packet routing feature. However, you can create one  
routing interface on the switches to serve as an IP configuration for the  
device. For more information, refer to “Routing Interfaces and  
ˆ The commands for managing interfaces are ADD IP INTERFACE,  
DELETE IP INTERFACE, and SET IP INTERFACE.  
Section VII: Routing  
303  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 27: Internet Protocol Version 4 Packet Routing  
Note  
Routing interfaces can be configured from either the command line  
interface or the menus interface.  
The following subsections describe the three main components of a  
routing interface:  
ˆ VLAN ID (VID)  
ˆ Interface number  
ˆ IP address and subnet mask  
VLAN ID (VID) An interface must be assigned to the VLAN on the switch where its  
network or subnet resides. The VLAN is identified by its VLAN  
identification (VID) number or VLAN name. The sequence of operations is  
to create the VLAN first and then the routing interface. Creating the  
interface before the VLAN is not permitted.  
A VLAN can have more than one interface in circumstances where a  
virtual LAN contains more than one subnet. The maximum number is  
sixteen routing interfaces per VLAN, making sixteen the maximum number  
of subnets you can have in an VLAN and still support packet routing on all  
of them.  
Interface An interface must be assigned an interface number in the range of 0 to 15.  
This range corresponds to the maximum number of interfaces permitted in  
Numbers  
a VLAN. Interfaces in different VLANs on the same switch can have the  
same interface number, but interfaces in the same VLAN must have  
different numbers.  
For instance, if a switch has four local subnets and each is in a different  
VLAN, all of the interfaces could have the same interface number, such as  
0. However, if two or more of the subnets reside in the same VLAN, the  
routing interfaces for the subnets in the VLAN must be assigned different  
interface numbers.  
Interfaces numbers are only used for interface identification when there is  
more than one subnet and routing interface in a VLAN. Consequently, the  
sequence in which the interface numbers are used is not important.  
IP Address and An interface must be a member of the local network or subnet where it will  
function as the logical connection for routing IPv4 packets. As such, it  
must be assigned a unique IP address and a subnet mask appropriate to  
the network or subnet.  
Subnet Mask  
The IP address and subnet mask of an interface can be assigned  
manually or supplied by a DHCP or BOOTP server on the network. When  
a VLAN contains more than one interface, only one of the interfaces can  
obtain its IP address from a DHCP or BOOTP server. The IP addresses of  
304  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
AT-S63 Management Software Features Guide  
the other interfaces in the same VLAN must be assigned manually. For  
example, if there are four interfaces and each of their respective subnets  
resided in a separate VLAN, then each interface can obtain its IP address  
and subnet mask from a DHCP or BOOTP server. However, if the four  
subnets share the same VLAN, only one interface can obtain its IP  
address from a DHCP or BOOTP server. The other three must be  
configured manually.  
Section VII: Routing  
305  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 27: Internet Protocol Version 4 Packet Routing  
Interface Names  
Many of the IPv4 routing commands have a parameter for an interface  
name. An interface name consists of a VLAN and an interface number,  
separated by a dash. The VLAN is designated by “vlan” followed by the  
VLAN identification number (VID) or the VLAN name.  
Here are several examples. The name for a interface in a VLAN with the  
VID of 7 and an interface number of 0 is:  
vlan7-0  
The name for an interface in a VLAN with the VID of 28 and an interface  
number of 2 is:  
vlan28-2  
Here is an example of an interface name that uses the VLAN name  
instead of the VID to identify the VLAN. The interface is part of the Sales  
VLAN and has an interface number of 5. Note that a dash separates “vlan”  
from the VLAN name.  
vlan-Sales-5  
The following is an example of a command that uses an interface name.  
The example uses the ADD IP INTERFACE command to create a new  
interface for a subnet in a VLAN with a VID of 28. The interface is  
assigned an interface number of 0, an IP address of 149.44.22.22, and a  
subnet mask of 255.255.255.0:  
add ip interface=vlan28-0 ipaddress=149.44.22.22 mask  
255.255.255.0  
This command is identical to the previous command, except it identifies  
the VLAN by its name, Production:  
add ip interface=vlan-Production-0 ipaddress=149.44.22.22  
mask 255.255.255.0  
306  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Static Routes  
In order for the switch to route an IPv4 packet to a remote network or  
subnet, there must be a route to the destination in the routing table of the  
switch. The route must consist of the IP address of the remote destination  
and the IP address of the next hop to reaching the destination.  
One type of route to a remote destination is referred to as a static route.  
You create static routes by manually entering them into the routing table.  
Static routes are never deleted from the routing table by the switch, even  
when they are inactive.  
When you create a static route, the switch’s management software  
automatically adds it to the interface that is a part of the same subnet as  
the next hop of the route. Consequently, before you can create a static  
route, the switch must have a routing interface that is a member of the  
same subnet as the next hop of the route.  
For example, assume a switch supported four subnets with four interfaces  
named VLAN4-0, VLAN11-0, VLAN12-0, and VLAN12-1. If you created a  
static route to a remote destination that had as its next hop an IP address  
in the subnet of the VLAN4-0 interface, the switch would automatically add  
the route to the VLAN4-0 interface.  
A new static route immediately becomes available for all of the interfaces  
on a switch to use for routing packets to the remote subnet. For example,  
referring to the previous example, a static route added to the VLAN4-0  
interface would be available to all the other interfaces on the same switch.  
The switch can store up to 1024 static routes.  
A static route is functional as soon as it is added to an interface and cannot  
be disabled. You must delete a route from the table to stop a switch from  
routing packets with the route.  
Static routes have a parameter called the metric that is a measurement of  
the cost of the switch when it forwards packets to the remote destination  
specified in the static route. The metric or cost is simply the hop count. The  
default setting for a static route is one hop. The value can be set higher to  
make a static route more costly. Networks, subnets, and nodes directly  
connected to a router have a hop count of 0.  
When the switch receives a packet from a remote subnet, it increases the  
metric or cost of the packet before forwarding it on to the next hop. A  
remote destination with a hop count of 16 is considered unreachable.  
Static routes also have a parameter for assigning a preference value. The  
switch uses this value to select the active routes when there are multiple  
static or dynamic routes in the routing table to the same remote  
Section VII: Routing  
307  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 27: Internet Protocol Version 4 Packet Routing  
destination. The range for the preference parameter is 0 to 65535. The  
lower the value, the higher the preference. The default value for a static  
route is 60.  
The commands for managing static routes are ADD IP ROUTE, DELETE  
IP ROUTE, and SET IP ROUTE.  
Note  
The command line interface is the only management interface in the  
AT-S63 Management Software that supports static routes. Static  
routes are not supported from the menus and web browser  
interfaces.  
308  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Routing Information Protocol (RIP)  
A switch can automatically learn routes to remote destinations by sharing  
the contents of its routing table with its neighboring routers in the network  
with the Routing Information Protocol (RIP) versions 1 and 2.  
RIP is a fairly simple distance vector routing protocol that defines networks  
based in how many hops they are from the switch, just as with static  
routes. Once a network is more than fifteen hops away (one hop is one  
link), it is considered as unreachable and is not included in the routing  
table.  
RIP version 2 permits the addition of subnet masks and next hop  
information in RIP updates. This allows the use of different sized subnet  
masks on different subnets within the same network.  
RIP broadcasts are automatically activated when the protocol is added to  
a routing interface on the switch. An interface sends RIP packets to the  
RIP multicast address 224.0.0.9 when sending version 2 packets or uses  
the broadcast address when sending out version 1 packets.  
A route is propagated by RIP if its status at the physical level is active. An  
active route has at least active one port in the VLAN. RIP does not  
propagate an inactive route where there are no active ports in the VLAN.  
RIP can be added to a maximum of 100 interfaces on a switch and the  
route table can store up to 1024 dynamic routes.  
Since the interfaces on a switch can route packets among the local  
subnets without the presence of RIP or static routes, the routing protocol is  
only necessary if the switch is to learn remote destinations by sharing the  
switch’s routing table with the neighboring routers, and you choose not to  
specify the routes manually with static routes.  
You add RIP to the routing interfaces where there are neighboring routers  
to remote destinations. You do not need to add RIP to interfaces where  
there are no neighboring routers.  
A route learned by RIP is immediately added to the routing table, where it  
becomes available to all the interfaces on the switch.  
When you add RIP to an interface, you can specify the type of RIP packets  
the routing protocol is to send and receive. The AT-9400 Switch can send  
either version 1 or 2 packets and accept either or both versions.  
Version 2 supports the addition of a password of up to sixteen  
alphanumeric characters to protect routers and their tables from  
incorporating bogus routing updates. The switch adds the password into  
the routing table when it broadcasts the contents of the table to its  
neighboring routing devices, which check the password prior to updating  
Section VII: Routing  
309  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 27: Internet Protocol Version 4 Packet Routing  
their tables.  
Note  
A RIP version 2 password is sent in plaintext. The AT-S63  
Management Software does not support encrypted RIP passwords.  
The switch broadcasts its routing table every thirty seconds from those  
interfaces that have RIP. This interval is not adjustable on the switch. The  
entire table is sent with the following exceptions:  
ˆ Dynamic RIP routes that fall under the split horizon rule.  
ˆ Inactive interface routes where there are no active ports in the VLAN.  
Note  
The AT-S63 Management Software does not support the RIP  
holddown and flush timers.  
The commands for managing RIP are ADD IP RIP, DELETE IP RIP, and  
SET IP RIP.  
Note  
RIP must be configured from the command line interface. The  
menus and web browser interfaces do not support this feature.  
The AT-9400 Switch supports the following RIP functions:  
ˆ Split horizon  
ˆ Split horizon with poison reverse  
ˆ Autosummarization of routes  
310  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Default Routes  
A default route is used when the switch cannot find a route in its routing  
table for a packet that needs to be forwarded to a remote destination.  
Rather than discard the packet, the switch sends it to the next hop  
specified in the default route.  
A default route has an destination IP address of 0.0.0.0 and no subnet  
mask. A default route can be enter manually in the form of a static route or  
learned dynamically through RIP. A switch can have multiple default  
routes.  
Section VII: Routing  
311  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 27: Internet Protocol Version 4 Packet Routing  
Equal-cost Multi-path (ECMP) Routing  
The routing table uses ECMP to store multiple routes to a remote  
destination so that the switch can distribute the traffic load over several  
routes. This can improve network performance by increasing the available  
bandwidth for traffic flows. It can also provide route redundancy.  
The routing table permits up to 32 routes to the same remote destination,  
with up to eight of the routes as active at one time. The routes can be all  
static routes, RIP routes, or a combination of the two. Routes to the same  
destination must have different next hops. The routing table will not permit  
two entries to the same remote destination with the same next hop.  
When the routing table contains eight or less routes to the same  
destination, all the routes can be active and available to route packets.  
The distribution of the traffic among the active routes is controlled through  
a hash that combines the packet source and destination IP addresses to  
select a route for packets from a source node. The traffic from a specific  
source and destined for a specific remote destination is assigned a route  
and all traffic to that remote destination from that source is forwarded  
using that route. The assignment of a route does not change except if the  
path is lost (for instance, the status of an interface changes from up to  
down), in which case its traffic is redirected to one of the remaining routes.  
When there are more than eight routes in the table to the same  
destination, the active routes are selected by preference value, metric  
value, and age, in that order. The routes with the eight lowest preference  
values are selected as the active routes. Where routes have the same  
preference value, selection is based on the lowest metric values.  
Otherwise, the selections are based on when the routes were added to the  
routing table, with older routes given preference over newer routes.  
Those routes not selected as active routes are placed in a standby mode.  
The selection of the active destination routes by the switch is dynamic and  
can change as routes are added and deleted from the routing table, and  
when they change status. For instance, if a new static or RIP route is  
added to the routing table when there are already eight active routes to the  
same destination, the new route will replace an existing active route if it  
has a lower preference value, forcing one of the active routes to change to  
the standby mode.  
Furthermore, an interface must be physically up with at least one active  
port in the VLAN for any of its routes to be considered as available for use.  
If an interface is down, meaning there are no active ports in the VLAN, the  
routes of the interface are considered inactive and are not assigned any  
traffic. For example, if there are eight routes to the same destination, but  
two of the routes reside in an interface that is down, those routes are not  
used, leaving six available routes.  
312  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
ECMP also applies to default routes. This enables the switch to store up to  
32 default routes with up to eight of the routes active at one time.  
The ECMP feature can be enabled and disabled on the switch. The  
operating status of ECMP does not affect the switch’s ability to store  
multiple routes to the same destination in its routing table. Rather, it  
controls how many of the available routes the switch uses to route packets  
to the same remote destination. When ECMP is enabled, the default  
setting, the switch can use multiple routes to route packets to the same  
remote destination, as explained above. When ECMP is disabled, the  
switch selects only one route, based on preference value, metric value, or  
age, to route packets to a remote subnet even when there are multiple  
routes to the subnet.  
A local subnet or directly connected network of a switch is usually  
represented just once in the routing table by its routing interface. However,  
in some situations a local subnet might have several routes to it if it is also  
remotely reachable through other routing interfaces on the switch via other  
routers. One of the routes would be the subnet’s routing interface and the  
others could be RIP or static routes. Here, even if ECMP is enabled, the  
switch uses only the routing interface to route packets to the local subnet,  
because the routing interface is always the preferred route to a local  
subnet. Any RIP or static routes to the local subnet are held in the standby  
mode for fail-over protection. They are only used when the status of the  
routing interface to the local subnet is down.  
Section VII: Routing  
313  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 27: Internet Protocol Version 4 Packet Routing  
Routing Table  
The switch maintains its routing information in a table of routes that tells  
the switch how to find a local or remote destination. Each route is uniquely  
identified in the table by its IP address, network mask, next hop, protocol,  
and routing interface.  
When the switch receives an IPv4 packet, it scans the routing table to find  
the most specific route to the destination on an “up” interface where there  
is at least one active port in the VLAN. If the switch does not find a direct  
route to the remote destination and no default route exists, the switch  
discards the packet and sends an ICMP message to that effect back to the  
source.  
The switch transmits its routing table every thirty seconds from those  
interfaces that have RIP. The RIP timer is not adjustable. The switch also  
transmits its routing table and resets the timer to zero whenever there is a  
change to the table to ensure that the neighboring routers are immediately  
informed of updates to the table.  
Dynamic RIP routes are removed from the table when they are not kept up  
to date (refreshed) by the neighboring routers. The metric of a route that is  
not refreshed is increased to 16 to indicate an unreachable network. If the  
route is not updated after 180 seconds, it is deleted from the table.  
The maximum storage capacity of the routing table in the AT-9400 Switch  
is:  
ˆ 512 interface routes  
ˆ 1024 static routes  
ˆ 1024 RIP routes  
314  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Address Resolution Protocol (ARP) Table  
The switch maintains an ARP table of IP addresses and the matching  
Ethernet MAC addresses. It refers to the table when routing packets to  
determine the destination MAC addresses of the nodes, as well as  
interfaces and ports from where the nodes are reached.  
The ARP table can store both static and dynamic entries. Static entries are  
entries you add yourself. This type of entry is never removed by the switch  
from the ARP table, even when the corresponding nodes are inactive.  
Dynamic entries are entries that the switch learns on it own. Dynamic  
entries of inactive nodes are periodically removed from table to prevent the  
table from filling with entries of inactive nodes.  
The switch learns addresses by sending out ARP requests. It generates  
an ARP request whenever it receives a packet that needs to be routed  
across a subnet, but lacks the destination MAC address in its ARP table.  
The switch, after receiving the ARP response from the destination node,  
adds the IP address and MAC address of the node to its ARP table and  
begins to route packets to the device. It should be noted that the switch  
discards all packets intended to a destination node until it receives a  
respond to its ARP request.  
The switch can also learn addresses when it is the destination of an ARP  
request from another node, such as when it is pinged by a management  
station. The switch adds the source IP address and MAC address in the  
request from the node to the table when it responds to the ARP request.  
Dynamic ARP entries are aged from the table according to the ARP cache  
timeout value to protect the table from filling with entries for hosts which  
are no longer active. The default setting for the timeout value is 150  
seconds. This value is adjustable with the SET IP ARP TIMEOUT  
command. Static ARP entries are not aged and are retained in the table  
even when the nodes are inactive.  
The commands for managing the ARP table are ADD IP ARP, DELETE IP  
ARP, SET IP ARP, SET IP ARP TIMEOUT, and SHOW IP ARP.  
Note  
The switch does not support Proxy ARP.  
The storage capacity of the ARP table in the AT-9400 Switch is:  
ˆ 1024 static entries  
ˆ 1024 dynamic entries  
Section VII: Routing  
315  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 27: Internet Protocol Version 4 Packet Routing  
Internet Control Message Protocol (ICMP)  
ICMP allows routers to send error and control messages to other routers  
or hosts. It provides the communication between IP software on one  
system and IP software on another. The switch implements the non-  
obsolete ICMP functions listed in Table 24.  
Table 24. ICMP Messages Implemented on the AT-9400 Switch  
ICMP Packet (Type)  
Echo reply (0)  
Switch Response  
This is used to implement the  
“ping” command common to most  
UNIX and TCP implementations.  
The switch sends out an “Echo  
reply” packet in response to an  
“Echo request.”  
Destination unreachable (3)  
Source Quench (4)  
This message is sent out when the  
switch drops a packet because it  
did not have a route to the  
destination.  
The switch will send a “Source  
Quench” if it must drop a packet  
due to limited internal resources.  
This could be because the source  
was sending data too fast to be  
forwarded.  
Redirect (5)  
The switch will issue a “redirect”  
packet to inform a local host that  
its target is located on the same  
LAN (no routing is required) or  
when it detects a host using a non-  
optimal route (usually because a  
link has failed or changed its  
status).  
Ech request (8)  
This is related to (1) and results in  
an “echo reply” packet being sent.  
The switch can also generate an  
“echo request” packet as a result  
of the PING command.  
316  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Table 24. ICMP Messages Implemented on the AT-9400 Switch  
ICMP Packet (Type)  
Switch Response  
Time to Live Exceeded (11)  
If the TTL field in a packet falls to  
zero the switch will send a “Time to  
live exceeded” packet. This could  
occur if a route was excessively  
long or if too many hops were in  
the path.  
Section VII: Routing  
317  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 27: Internet Protocol Version 4 Packet Routing  
Routing Interfaces and Management Features  
Routing interfaces are primary intended for the IPv4 packet routing  
feature. There are, however, a number of management functions that rely  
on the presence of at least one routing interface on the switch to operate  
properly. The switch uses the IP address of an interface as its source  
address when performing the management function. The management  
functions are listed here:  
ˆ Network servers  
ˆ Enhanced stacking  
ˆ Remote Telnet, SSH, and web browser management sessions  
ˆ Pinging a remote device  
ˆ DHCP or BOOTP server  
Network Servers A local subnet on the switch must have an interface if the device is using  
the subnet to access any of the following types of network servers:  
ˆ SNTP server for setting the switch’s date and time.  
ˆ RADIUS or TACACS+ authentication server for manager access  
accounts and 802.1x port-based network access control.  
ˆ Syslog server for storing events from the switch’s event logs.  
ˆ TFTP server for uploading and downloading files to the switch.  
The switch uses the IP address of the interface as its source address  
when communicating with the network server. Without a routing interface  
on the subnet, the switch will not have a source IP address to include in its  
packets. For example, in order to set its date and time using an SNTP  
server, the switch must have a routing interface on the local subnet from  
where it is reaching the server.  
The servers can be located on different routing interfaces on the switch.  
For instance, the switch can access an SNTP server through one interface  
and a RADIUS authentication server from another. This differs from some  
of the earlier versions of the AT-S63 Management Software where all the  
servers had to be members of what was referred to as the “management  
VLAN.”  
If you intend to use the IPv4 routing feature of the switch and assign  
routing interfaces to all the local subnets and networks on a switch, this  
requirement should not be a issue. However, if you choose not to use the  
routing function and so not create interfaces or you have an AT-9400  
Switch that only supports one interface, some planning will be necessary  
in order to use these features. At a minimum, you must create one routing  
interface on the switch and plan your network so that the switch can  
access the servers from the subnet of the interface.  
318  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
AT-S63 Management Software Features Guide  
As an example, assume you decided not to implement the IPv4 routing  
feature on a switch that had four local subnets, but you wanted the switch  
to send its events to a syslog server and have access to a RADIUS  
authentication server. Assume also that you wanted to use a TFTP server  
to upload and download files to the device. This would require that you  
plan your network so that the switch could reach the syslog, RADIUS, and  
TFTP servers from the same local subnet on the unit. You would also need  
to assign an interface to the subnet. The switch, having only one interface,  
would not route IPv4 packets among its local subnets and directly  
connected networks, but would use the interface’s IP address to  
communicate with the servers.  
Enhanced The enhanced stacking feature simplifies the task of managing the Allied  
Telesis switches in your network by allowing you to easily transition among  
the switches in a stack during a management session.  
Stacking  
The switches of an enhanced stack must be interconnected by a common  
VLAN and the VLAN must be assigned a routing interface on each of the  
switches. Furthermore, the routing interface in the common VLAN on the  
master switch must be designated as the local interface, as described in  
There is an important difference between the need for interfaces with  
enhanced stacking versus network servers, as explained in the previous  
subsection. Network servers can be reached by the switch through  
different interfaces in different subnets, simultaneously. In contrast, the  
switches of an enhanced stack must share a common VLAN and subnet.  
For background information and guidelines on the enhanced stacking  
Remote Telnet, Remote Telnet or SSH management sessions of a switch must be  
transacted through a subnet that has been assigned a routing interface on  
the unit. Furthermore, the interface must be designated as the switch’s  
local interface. Only a workstation that can reach the switch through the  
SSH, and Web  
Browser  
Management  
Sessions  
subnet of the local interface can manage the unit. This rule applies to  
isolated devices (that is, switches that are not a part of an enhanced stack)  
and a master switch of an enhanced stack. This does not apply to a slave  
switches of an enhanced stack.  
For background information and guidelines on remote management, refer  
to the Starting an AT-S63 Management Session Guide.  
Section VII: Routing  
319  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 27: Internet Protocol Version 4 Packet Routing  
Pinging a Remote This function is used to validate the existence of an active path between  
the switch and another network node. The switch can ping a device if  
Device  
there is a routing interface on the local subnet from where it reaches the  
device. In previous versions of the AT-S63 Management Software the  
device to be pinged had to be reached through the management VLAN of  
the switch. This restriction no longer applies. A remote device can be  
pinged from any subnet of the switch that has an interface.  
DHCP or You can use a DHCP or BOOTP server to assign IP addresses to the  
interfaces of a switch by activating the client on the interfaces.  
BOOTP Server  
320  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Local Interface  
The local interface is used with the enhanced stacking feature. It is also  
used with remote management of a switch with a Telnet or SSH client, or a  
web browser. The local interface does the following:  
ˆ With an enhanced stack, it designates on the master switch the  
common VLAN and subnet that interconnects the switches of the  
stack. The master switch uses the local interface to send out its  
broadcast packets when searching for the other switches in a stack.  
ˆ With remote management, it designates the VLAN and subnet from  
where the remote management workstation will access the switch. The  
switch uses the local interface to watch for the management packets  
from the remote workstation and to send packets back to the remote  
station.  
For example, assume you wanted to remotely manage a switch that had  
four subnets and four interfaces named VLAN4-0, VLAN11-0, VLAN12-0,  
and VLAN12-1, and the remote workstation was reaching the switch  
through the subnet of the VLAN11-0 interface. You would need to  
designate the VLAN11-0 interface as the local interface on the unit.  
A switch can have only one local interface.  
For background information on remote management of the switch, refer to  
the Starting an AT-S63 Management Session Guide. For background  
information on enhanced stacking, refer to Chapter 2, “Enhanced  
Section VII: Routing  
321  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 27: Internet Protocol Version 4 Packet Routing  
AT-9408LC/SP AT-9424T/GB, and AT-9424T/SP Switches  
The AT-9408LC/SP, AT-9424T/GB, and AT-9424T/SP Switches do not  
support the IPv4 packet routing feature. They do, however, support a  
limited version of some of the features.  
Local Interface You can create one routing interface to provide support for those  
management features that require the switch to have an IP address.  
Furthermore, the interface can be designated as the local interface so that  
the switch can function as the master switch of an enhanced stack or for  
remote Telnet, SSH, or web browser management. For further  
ARP Table These switches also have an ARP table with a maximum capacity of ten  
ARP entries. The table and entries are used by the AT-S63 Management  
Software when it performs a management function that requires it to  
communicate with another device on the network. An example would be if  
you instructed the switch to ping another network device or download a  
new AT-S63 image file or configuration file from a TFTP server.  
The value of the ARP table is that it eliminates the need of the switch to  
issue unnecessary ARP broadcast packets when performing some  
management functions. This can improve the switch’s response time as  
well as reduce the number of broadcast packets on your network.  
There are two types of entries. One type is permanent. There is only one  
permanent entry and it is used by the switch for internal diagnostics. It can  
never be removed from the table.  
The other type is a temporary entry, of which there can be up to nine. The  
switch adds a temporary entry whenever its management software  
interacts with another network device during a management function.  
When you enter a management command that contains an IP address not  
in the table, the switch sends out an ARP broadcast packet. When the  
remote device responds with its MAC address, the switch adds the  
device’s IP address and MAC address as a new temporary entry to the  
table.  
A temporary entry remains in the table only while active. An entry remains  
active so long as it is periodically used by the switch for management  
functions. If an entry is inactive for a defined period of time known as the  
ARP cache timeout, it is automatically removed from the table. To adjust  
this value, refer to the SET IP ARP TIMEOUT command. The default is  
150 seconds. If the table becomes full, the management software  
continues to add new entries by deleting the oldest entries.  
322  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Note  
The AT-9408LC/SP, AT-9424T/GB, and AT-9424T/SP Switches do  
not use the ARP table to move packets through the switching matrix.  
They refer to the table only when they perform a management  
function requiring them to communicate with another network node.  
Default Gateway The default gateway specifies the IP address of an interface on a  
neighboring router. The switch’s management software uses this address  
as the next hop to reaching a remote network device, such as a remote  
Telnet, SSH, or web browser management workstation or a syslog server,  
when the switch’s interface and the remote device are on different  
subnets.  
As an example, assume you wanted to manage the switch from a remote  
management workstation on a different subnet than the local interface,  
and needed the switch to access a RADIUS authentication server also on  
a different subnet. Here, you would need to define a default gateway on  
the switch so that the unit would know the next hop to reaching the remote  
workstation and the RADIUS server.  
The default gateway is only used for management functions, such as  
communicating with a remote management workstation or sending events  
to a syslog server. The default gateway is not used during the normal  
Layer 2 switching of packets among the switch ports and, as such, is not  
necessary for normal operations of the device.  
You define the default gateway by creating a default route on the switch.  
As explained in “Static Routes” on page 307, this type of route does not  
specify a destination address. Rather, it simply defines the IP address of  
the next hop, which becomes the default gateway for the switch.  
The IP address of the next hop of the default route must be of the same  
subnet as the switch’s interface.  
Section VII: Routing  
323  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 27: Internet Protocol Version 4 Packet Routing  
Routing Command Example  
This section contains an example of the IPv4 routing feature. It illustrates  
the sequence of commands to implementing the feature. To make the  
example easier to explain, some of the command options are not  
mentioned and the default values are used instead. For information on all  
of the available options of a command, refer to the appropriate section in  
this chapter.  
Note  
This example does not apply to the AT-9408LC/SP, AT-9424T/GB,  
and AT-9424T/SP switches, which do not support the packet routing  
feature. For an example of how to assign an IP address to these  
This example has the following sections:  
This example assumes an AT-9448T/SP switch with four local subnets.  
Two subnets will reside in their own VLANs and two will share a VLAN.  
The table below lists the relevant information.  
Table 25. IPv4 Routing Example  
Company  
Department  
Subnet IP  
Address  
Portsa  
VLAN Name  
VID  
Subnet Mask  
Sales  
Sales  
4
5
149.35.67.0  
255.255.255.0 U - 1-11  
T- 50  
Production  
Production  
149.35.68.0  
149.35.69.0  
149.35.70.0  
255.255.255.0 U - 12-20  
T - 50  
Engineering  
Group 1  
Engineering  
11  
255.255.255.0 U - 21 - 40  
T - 50  
Engineering  
Group 2  
255.255.255.0  
a. U - untagged ports; T - tagged ports  
324  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Creating the The first step is to create the VLANs for the local subnets on the switch.  
The VLANs must be created before the routing interfaces. The following  
command creates a VLAN for the Sales department with a VID of 4 and  
the appropriate ports:  
VLANs  
create vlan=Sales vid=4 untaggedport=1-11 taggedport=50  
The following commands create the Production and Engineering VLANs:  
create vlan=Production vid=5 untaggedport=12-20  
taggedport=50  
create vlan=Engineering vid=11 untaggedport=21-40  
taggedport=50  
Note that even though there are four local subnets in the example, there  
are only three VLANs because two of the subnets will share a VLAN.  
For further information on this command, refer to the CREATE VLAN  
command.  
Creating the Now that the VLANs are created, you can add the routing interfaces for the  
individual subnets. There are four local subnets in the example, so there  
will need to be four interfaces to support routing on all of them.  
Routing  
Interfaces  
The following command creates the routing interface for the Sales subnet.  
The interface name is based on the VID of the VLAN, which is 4, and an  
interface number, in this case 0. The interface is assigned the unique IP  
address 149.35.67.11 and a subnet mask to make it a member of its  
corresponding subnet.  
add ip interface=vlan4-0 ipaddress=149.35.67.11  
netmask=255.255.255.0  
These commands create the interfaces for the remaining subnets:  
add ip interface=vlan5-0 ipaddress=149.35.68.24  
netmask=255.255.255.0  
add ip interface=vlan11-0 ipaddress=149.35.69.23  
netmask=255.255.255.0  
add ip interface=vlan11-1 ipaddress=149.35.70.45  
netmask=255.255.255.0  
The Engineering VLAN (VID 11) has two interfaces for its two subnets.  
Each interface is given a different interface number, 0 and 1, to distinguish  
between them.  
At this point, the switch begins to route IPv4 packets among the local  
subnets.  
For further information on this command, refer to the ADD IP INTERFACE  
Section VII: Routing  
325  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 27: Internet Protocol Version 4 Packet Routing  
command.  
Adding a Static Building on our example, assume you decided to manually enter a route to  
a remote subnet as a static route. The command for creating a static route  
is ADD IP ROUTE. Here is the basic information for defining a static route:  
Route and  
Default Route  
ˆ The IP address of the remote destination.  
ˆ The subnet mask of the remote destination.  
ˆ The IP address of the next hop.  
ˆ The routing interface on the switch where the next hop is located. This  
piece of information is optional because the switch can automatically  
determine the appropriate interface from the IP address of the next  
hop. The IP addresses of the next hop of a static route and the  
interface where the hop is located must be members of the same  
subnet.  
Let’s assume you wanted to add a static route to a remote subnet with the  
IP address 149.35.22.0 and a mask of 255.255.255.0. Let’s also assume  
that the IP address of the next hop is 149.35.70.26, making it part of the  
subnet of the VLAN11-1 interface. Consequently, the static route must be  
added to that interface, though you do not need to specify it in the  
command. Here is the command for adding the static route:  
add ip route=149.35.22.0 nexthop=149.35.70.26  
mask=255.255.255.0  
A static route becomes active as soon as it is defined and is available to all  
of the interfaces on the switch.  
Now assume that you want to create a default route for when the switch  
receives a packet with a destination address to a network or subnet for  
which it does not have a route. All you need to know for a default route is  
the IP address of the next hop for the packets. For this example, assume  
that the IP address of the next hop will be 149.35.68.12. This locates the  
next hop on the VLAN5-0 interface. Here is the command for creating the  
default route:  
add ip route=0.0.0.0 nexthop=149.35.68.12  
A default route does not have a subnet mask. Note also that the  
appropriate routing interface for the next hop, in this example VLAN5-0, is  
also not defined because, as with other static routes, specifying the  
interface is optional.  
326  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Adding RIP Rather than adding the static routes to remote destinations, or perhaps to  
augment them, you decide that the switch should learn routes by  
exchanging its route table with its routing neighbors using RIP. To  
implement RIP, you add it to the routing interfaces where routing  
neighbors are located. The command for adding RIP to an interface is  
ADD IP RIP.  
For the purpose of this example, assume the routing neighbors of the  
switch are located on the VLAN5-0 and VLAN11-1 interfaces. The  
following commands add RIP to the interfaces and configure the routing  
protocol to send only version 2 packets, but accept packets of either  
version 1 or 2. In both cases, RIP is running without a password.  
add ip rip interface=vlan5-0 send=rip2 receive=both  
authentication=none  
add ip rip interface=vlan11-1 send=rip2 receive=both  
authentication=none  
You could, if you wanted, add RIP to the other interfaces. But since, in our  
example, those interfaces do not have links to other RIP routers, they  
would not learn any routes.  
Selecting the This last part of the example designates a local interface. This step is  
necessary on a master switch of an enhanced stack to designate the  
Local Interface  
common VLAN of the switches in the stack. This is also necessary if you  
want to manage the device from a remote management workstation with a  
Telnet or SSH client, or a web browser.  
Let’s assume you plan to remotely manage the switch from a management  
workstation that reaches the device through the subnet in the Sales VLAN,  
which has the interface name is VLAN4-0. Here is the command to  
designate that interface as the local interface on the switch:  
set ip local interface=vlan4-0  
To start a remote management session on the switch, you use the IP  
address of the local interface as the switch’s address. In the example, the  
switch’s address would be 149.35.67.11 because that happens to be the  
IP address of the VLAN4-0 interface, which is the local interface.  
Section VII: Routing  
327  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 27: Internet Protocol Version 4 Packet Routing  
Non-routing Command Example  
This example illustrates how to assign an IP address to a switch by  
creating just one interface. This example is appropriate in cases where  
you want to implement the management functions described in “Routing  
packet routing. This section is also appropriate for the AT-9400 Layer 2+  
Switches, which do not support packet routing.  
The first step is to select the VLAN and subnet on the switch for the  
interface. The appropriate VLAN for the master switch of an enhanced  
stack is the common VLAN of the switches in the stack. The appropriate  
VLAN for remote management or for remote access to a network server is  
the VLAN where the remote device is located.  
Let’s assume for the purposes of this example that the switch will be  
remotely managed from a Telnet, SSH, or web browser management  
workstation on the network. Consequently, the appropriate VLAN would  
be the VLAN on the switch where the remote management workstation is  
located. Assume that the VID of the VLAN is 12 and the IP address of the  
subnet of the VLAN is 149.44.55.0 with a subnet mask of 255.255.255.0.  
The following command assigns an interface to the VLAN. It identifies the  
VLAN by its VID of 12 and assigns it the interface number 0. The interface  
is given the IP address 149.44.55.22 to make it a member of the subnet:  
add ip interface=vlan12-0 ipaddress=149.44.55.22  
netmask=255.255.255.0  
In order to manage the switch remotely, the interface must be designated  
as the local interface so that the management software monitors the  
subnet for management packets. Here is the command for designating the  
interface as the local interface:  
set ip local interface=vlan12-0  
As the final part of the example, assume that the management software on  
the switch must communicate with a network device, such as  
management workstation, syslog server, or RADIUS server, that is not a  
member of the same subnet as the interface. For this, you need to define a  
default route. The route will specify the next hop to reaching the remote  
subnet. The switch will use the default route whenever it needs to send a  
management packet to a remote network device that resides on a different  
subnet than its local interface.  
The next hop in the route must specify the IP address of a routing interface  
on a router in the network. Furthermore, the IP address of the routing  
interface must be a member of the same subnet as the interface on the  
switch.  
328  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
The following command creates a default route for the example and  
specifies the next hop as 149.44.55.6:  
add ip route=0.0.0.0 nexthop=149.44.55.6  
Section VII: Routing  
329  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 27: Internet Protocol Version 4 Packet Routing  
Upgrading from AT-S63 Version 1.3.0 or Earlier  
When the AT-9400 Switch running AT-S63 version 1.3.0 or earlier is  
upgraded to the latest version of the management software, the switch  
automatically creates a routing interface that preserves the previous IP  
configuration of the unit. If the switch had a static address, the interface is  
assigned the same address. If the address was supplied by a DHCP or  
BOOTP server, the DHCP or BOOTP client on the interface is activated.  
The interface is given the interface number 0 and assigned to the  
preexisting management VLAN. Furthermore, the interface is designated  
as the local interface of the switch.  
For example, a switch with the static IP address 149.55.55.55, subnet  
mask 255.2552.255.0, and a management VLAN with a VID of 12 will  
have, after the upgrade, a routing interface with the name VLAN12-0 and  
the same static IP address and subnet mask.  
By retaining the IP configuration, those management functions (e.g.,  
remote Telnet management, syslog client, and RADIUS client) that were  
operating before the upgrade will continue to operate after the unit is  
upgraded to the newest version of the management software. Without this  
feature, you would have to restore the switch’s IP configuration by  
manually creating a routing interface.  
If the switch does not have an IP address and the DHCP and BOOTP  
clients are not activated, the upgrade process does not create a routing  
interface.  
330  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 28  
BOOTP Relay Agent  
This chapter has the following sections:  
331  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 28: BOOTP Relay Agent  
Supported Platforms  
This feature is supported on the following switches:  
ˆ Layer 2+ Models  
Not supported  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature must be configured from the command line interface.  
332  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
The AT-S63 Management Software comes with a BOOTP relay agent for  
relaying BOOTP messages between clients and DHCP or BOOTP  
servers.  
When a client sends a BOOTP request to a DHCP or BOOTP server for an  
IP configuration, it transmits the request as a broadcast packet because it  
does not know the IP address of the server. This can present a problem  
when a client and server reside on different subnets, because broadcast  
packets do not cross subnet boundaries. One possible solution is to have  
a DHCP or BOOTP server on each subnet where there are clients, though  
this could be problematic if there happen to be a lot of subnets. Another  
solution is to use a BOOTP relay agent, which transfers client requests  
across subnet boundaries.  
The BOOTP relay agent does more than simply forward BOOTP requests  
from clients to servers. It modifies the requests so that, from the  
perspective of the server, it becomes the originator of the request. The  
responses from the servers are directed to the agent, which sends the  
messages on to the clients as either broadcast or unicast packets,  
depending on the requirements of the clients.  
To implement BOOTP relay on the switch, you need to be familiar with  
routing interfaces, which route packets between different local subnets on  
the switch in the IPv4 packet routing feature. Each routing interface  
functions as the BOOTP relay agent for the clients in its subnet, forwarding  
BOOTP requests from the clients and responses from the servers.  
If you will be using the IPv4 packet routing feature on all the local subnets,  
then, by default, all of the clients will have access to a BOOTP relay agent  
because each subnet will have a routing interface. However, if IPv4 packet  
routing will be limited to some but not all the local subnets of the switch,  
then only those BOOTP requests from clients on a subnet with a routing  
interface will be forwarded to a BOOTP relay agent.  
Here is an overview of the process. When a routing interface receives a  
BOOTP request with a value of 0.0.0.0 in the gateway (giaddr) field in the  
packet, it assumes the request originated from a client on its subnet. In  
response, it replaces the value in the field with its IP address and forwards  
the packet on to the server. If more than one IP address of DHCP or  
BOOTP servers are specified on the switch, the interface sends the same  
request to each server. If the client and server reside on the same subnet,  
the routing interface does not forward the request.  
If an interface receives a BOOTP request with a non-zero value in the  
gateway field, it assumes the client who originated the request resides on  
another subnet, and so routes the request as a unicast packet without any  
change, other than incrementing the hop count.  
Section VII: Routing  
333  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 28: BOOTP Relay Agent  
A routing interface that receives a BOOTP reply from a server inspects the  
broadcast flag field in the packet to determine whether the client, in its  
original request to the server, set this flag to signal that the response must  
be sent as a broadcast datagram. Some older nodes have this  
dependency. If the flag is not set, the routing interface forwards the packet  
to the originating client as a unicast packet. If the flag is set, the packet is  
forwarded as a broadcast by the interface.  
You configure the BOOTP relay agent on the switch by specifying the IP  
address of the BOOTP server on your network with the ADD BOOTP  
RELAY command. You can enter up to eight BOOTP or DHCP servers.  
The IP addresses apply to all the routing interfaces on the switch. BOOTP  
requests are forwarded to all the specified servers, simultaneously.  
You activate the BOOTP relay agent on the switch with the ENABLE  
BOOTP RELAY command. As soon as the agent is enabled the routing  
interfaces begin to forward BOOTP requests from the clients. Activating  
the client applies to all routing interfaces on the switch. You cannot  
activate the agent on some interfaces and not on others. The default  
setting for the agent on the switch is disabled.  
To view the status of the agent and the IP addresses of the servers, use  
the SHOW BOOTP RELAY command.  
334  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Guidelines  
These guidelines apply to the BOOTP relay agent:  
ˆ A routing interface functions as the BOOTP relay agent for the local  
clients in its subnet.  
ˆ You can specify up to eight DHCP or BOOTP servers.  
ˆ The hop count for BOOTP requests is preset on the AT-9400 Switch to  
4. It cannot be changed. Routing interfaces discard BOOTP requests  
with hop counts of 4 or more.  
ˆ Because both BOOTP and DHCP use BOOTP messages, the BOOTP  
relay agents can relay both their packets.  
Section VII: Routing  
335  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 28: BOOTP Relay Agent  
336  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 29  
Virtual Router Redundancy Protocol  
The chapter has the following sections:  
337  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 29: Virtual Router Redundancy Protocol  
Supported Platforms  
This feature is supported on the following switches:  
ˆ Layer 2+ Models  
Not supported  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature must be configured from the command line interface.  
338  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
This chapter describes the Virtual Router Redundancy Protocol (VRRP)  
support provided by the switch.  
One of the functions performed by switches is to act as a gateway to the  
WAN for hosts on a LAN. On larger LANs, two or more switches may act  
as the gateway, and hosts use a dynamic routing protocol, such as RIP or  
OSPF, to determine the gateway switch to use as the next hop in order to  
reach a specific IP destination. However, there are a number of factors,  
such as administrative or processing overhead, that may make it  
undesirable to use a dynamic routing protocol. One alternative is to use  
static routing. However, if the statically configured first hop switch fails, the  
hosts on the LAN are unable to communicate with hosts on the WAN.  
The Virtual Router Redundancy Protocol provides a solution to the  
problem by combining two or more physical switches into a logical  
grouping called a virtual router (VR). The physical switches in the virtual  
router operate together to provide a single logical gateway for hosts on the  
LAN.  
Section VII: Routing  
339  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 29: Virtual Router Redundancy Protocol  
Master Switch  
The virtual router has a virtual MAC address known by all the switches  
participating in the virtual router. The virtual MAC address is derived from  
the virtual router identifier, which is a user-defined value from 1 to 255. All  
hosts on the LAN are configured with an IP address to use as the first hop.  
This IP address is typically owned by the preferred switch in the group of  
switches that constitute the virtual router. When available, this switch  
performs the duties of the virtual router, and is referred to as the master.  
The switch that owns the IP address associated with the virtual router is  
referred to as the preferred master. When a virtual router is configured so  
that none of the participating switches owns the IP address, the virtual  
router has no preferred master.  
When a switch takes the role of master for a virtual router, it does the  
following:  
ˆ
Responds to ARP packets for the IP addresses associated with the virtual  
router. The ARP response contains the virtual MAC address of the virtual  
router so that the hosts on the LAN associate the virtual MAC address with  
their configured first hop IP address.  
ˆ
ˆ
ˆ
Forwards packets with a destination link layer MAC address equal to the  
virtual router MAC address.  
Accepts packets addressed to the IP address(es) associated with the virtual  
router, but only if it actually owns the address(es).  
Broadcasts advertisement packets at regular intervals (at the specified  
advertisement interval) to inform backup switches that it is still acting as the  
master switch.  
In accordance with the RFC standard, a user does not receive a response  
to ping or Telnet packets sent to the VR address unless the switch owns  
this address.  
340  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Backup Switches  
All the other switches participating in the virtual router are designated as  
backup switches. A switch can be part of several different virtual routers  
on one LAN, provided that all the virtual routers have different virtual router  
identifiers.  
When a switch functions as a backup for a virtual router, it does the  
following:  
ˆ Receives advertisement packets from the master switch and checks  
that the information contained in them is consistent with their own  
configuration, ignoring and discarding advertisement packets that do  
not match.  
ˆ Assumes the role of master switch for the virtual router if an  
advertisement packet is not received for a given period of time (the  
“master-down” period), based on the specified advertisement interval.  
The “master-down” time is approximately three times the  
advertisement interval.  
ˆ Assumes the role of master switch if it receives an advertisement  
packet from another switch with a lower priority than its own, if preempt  
mode is on.  
When the master switch fails, the backup switch assumes control and  
starts processing traffic.  
If a backup switch is about to assume the role of master of the virtual  
router because it has not received an advertisement for the “master-down”  
period, it first checks the operational status of the interface to which the  
virtual router is attached. If the interface is down, it does not enter the  
master state. Instead, it says in the backup state and checks the interface  
again after another “master-down” period, assuming that it does not  
receive an advertisement during that time.  
Section VII: Routing  
341  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 29: Virtual Router Redundancy Protocol  
Interface Monitoring  
The virtual router can monitor certain interfaces to change the priority of  
switches if the master switch loses its connection to the outside world.  
This is known as interface monitoring. Interface monitoring reduces the  
priority of the switch when an important interface connection is lost. The  
reduction in priority causes a backup switch with a higher priority to take  
over as the master switch and restore connectivity.  
If a master switch loses its connection to the outside world, the connection  
to the LAN is not affected. Advertisement packets are still sent by the  
master and received by the backup switches, but the master is unable to  
send data to other networks because its connection to the outside world  
has been lost  
342  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Port Monitoring  
Port monitoring is the process of detecting the failure of ports that are part  
of a VLAN that a virtual router is running over. If a port fails or is disabled,  
the VRRP priority is reduced by the stepvalue or by an amount that reflects  
the proportion of the VLAN’s ports that are out of service. If the switch is  
the master and a backup switch has a higher priority, the backup switch  
preempts the master and becomes the new master.  
Note the following about port monitoring:  
ˆ You can delete an IP interface if it is a monitored interface, because  
VRRP is only monitoring the state of the interface and does not require  
that the interface have an IP address.  
ˆ A VLAN cannot be destroyed if it is a monitored interface of a VRRP.  
To destroy a VLAN, you must first destroy the monitored interface.  
Section VII: Routing  
343  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 29: Virtual Router Redundancy Protocol  
VRRP on the Switch  
VRRP is disabled by default. When a virtual router is created on the  
switch, it is enabled by default, but the VRRP module must be enabled  
before it is operational. The VRRP module or a specific virtual router can  
be enabled or disabled afterwards by using the ENABLE VRRP and  
DISABLE VRRP commands.  
A virtual router must be created on at least two switches before it operates  
correctly. To create a virtual router for an IP address over an Ethernet  
interface, so that the switch participates in the virtual router, use the  
CREATE VRRP command.  
To destroy a virtual router on the LAN, it must be removed from all  
participating switches. To remove a virtual router so that the switch no  
longer participates in it, use the DESTROY VRRP command.  
If the switch in the master role for the virtual router becomes unavailable,  
the master role is taken by the switch with the highest priority amongst the  
available switches. The priority is a value from 1 to 255, with a default of  
100. The highest value of 255 is reserved for the switch that owns the  
virtual router’s IP address. The new master takes over all the  
responsibilities of the original master. Hosts on the LAN can continue  
sending packets to the same virtual MAC address with which they  
associate the configured first hop IP address, even though the switch that  
owns the IP address is not currently available. When the preferred switch  
that owns the IP address becomes available again, it resumes the role of  
master.  
By default, when a switch becomes available with a higher priority than the  
master, it takes over as master. This is referred to as preempt mode and  
can be set on or off. Even with preempt mode off, the switch that owns the  
IP address always becomes the master when available. If two switches  
are configured with the same priority, the one with the highest IP address  
has higher priority. Preempt mode must be the same for all switches in the  
virtual router. Set the priority and preempt mode when you create the  
virtual router; modify it later by using the SET VRRP command.  
The frequency with which the master sends advertisement packets must  
be set to the same value for all switches in the virtual router. The default  
advertisement interval of 1 second is recommended for most networks.  
This is set with the ADINTERVAL parameter in the CREATE VRRP and  
SET VRRP commands.  
Each of the switches in the virtual router can be configured for plaintext  
authentication or none. No authentication is suitable when there is minimal  
security risk, and the configuration is so simple (for example, two switches  
on a LAN) that there is little chance of configuration errors. Plaintext  
password authentication protects against accidental misconfiguration and  
344  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
prevents a switch from inadvertently backing up another switch. The  
authentication type and, in the case of plaintext authentication, the  
password, must be the same for all switches in the virtual router. By  
default, the virtual router has no authentication. Authentication is set with  
the AUTHENTICATION and PASSWORD parameters in the CREATE  
VRRP and SET VRRP commands.  
In order for the security level of the LAN to be maintained, each switch in  
the virtual router must have at least the minimum allowable level of  
security.  
A virtual router is always created to back up one primary IP address on all  
the switches in the virtual router. Up to 16 secondary IP addresses can be  
backed up by the same virtual router, as long as they are compatible with  
the IP address and mask associated with the Ethernet interface over which  
the IP address of the virtual router is operating. Such secondary  
addresses must be added to all the switches in the virtual router. The  
virtual router’s primary IP address cannot be deleted. To add or remove  
secondary IP addresses, use the ADD VRRP IPADDRESS and DELETE  
VRRP IPADDRESS commands.  
A monitored interface is one that the virtual router is dependent on for full  
operation. VRRP is informed if the operational status of the interface  
changes. If the interface is not operational, the switch’s priority is reduced.  
To add or remove a monitored interface to or from a virtual router, use the  
ADD VRRP MONITOREDINTERFACE and DELETE VRRP  
MONITOREDINTERFACE commands.  
It is important that all switches involved in a virtual router be configured  
with the same values for the following:  
ˆ VRRP virtual router identifier  
ˆ IP address  
ˆ advertisement interval  
ˆ preempt mode  
ˆ authentication type  
ˆ password  
Inconsistent configuration causes advertisement packets to be rejected  
and the virtual router cannot perform properly.  
Section VII: Routing  
345  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 29: Virtual Router Redundancy Protocol  
346  
Section VII: Routing  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Section VIII  
Port Security  
The chapters in this section contain overview information on the port  
security features of the AT-9400 Switch. The chapters include:  
Section VIII: Port Security  
347  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
348  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 30  
MAC Address-based Port Security  
The sections in this chapter include:  
Section VIII: Port Security  
349  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 30: MAC Address-based Port Security  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
Note  
This port security feature is not supported on GBIC, SFP, or XFP  
modules.  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
350  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Overview  
You can use this feature to enhance the security of your network by  
controlling which end nodes can forward frames through the switch, and  
so prevent unauthorized individuals from accessing your network. It uses a  
frame’s source MAC address to determine whether the switch should  
forward a frame or discard it. The source address is the MAC address of  
the end node that sent the frame.  
There are four levels of port security:  
ˆ Automatic  
ˆ Limited  
ˆ Secured  
ˆ Locked  
You set port security on a per port basis. Only one security level can be  
active on a port at a time.  
Automatic The Automatic security mode disables port security on a port. This is the  
default security level for a port.  
Limited The Limited security level allows you to specify the maximum number of  
dynamic MAC addresses a port can learn. The port forwards only packets  
of learned source MAC addresses and discards ingress frames with  
unknown source MAC addresses.  
When the Limited security mode is initially activated on a port, all dynamic  
MAC addresses learned by the port are deleted from the MAC address  
table. The port then begins to learn new addresses, up to the maximum  
allowed. After the port has learned its maximum number of addresses, it  
does not learn any new addresses, even when end nodes are inactive.  
A dynamic MAC address learned on a port operating in the Limited  
security mode never times out from the MAC address table, even when  
the corresponding end node is inactive.  
Static MAC addresses are retained by the port and are not included in the  
count of maximum dynamic addresses. You can continue to add static  
MAC addresses to a port operating with this security level, even after the  
port has already learned its maximum number of dynamic MAC  
addresses. A switch port can have up to 255 dynamic and static MAC  
addresses.  
Section VIII: Port Security  
351  
Download from Www.Somanuals.com. All Manuals Search And Download.  
           
Chapter 30: MAC Address-based Port Security  
Secured This security level uses only static MAC addresses assigned to a port to  
forward frames. Consequently, only those end nodes whose MAC  
addresses are entered as static addresses are able to forward frames  
through a port. Dynamic MAC addresses already learned on a port are  
discarded from the MAC table and no new dynamic addresses are added.  
Any ingress frames having a source MAC address not entered as a static  
address on a port are discarded.  
After activating this security level, you must enter the static MAC  
addresses of the end nodes that are to forward frames through the port.  
Locked A port set to this security level immediately stops learning new dynamic  
MAC addresses and forwards frames using the dynamic MAC addresses  
it has already learned and any static MAC addresses assigned to it.  
Ingress frames with an unknown MAC address are discarded. Dynamic  
MAC addresses already learned by a port prior to the activation of this  
security level never time out from the MAC address table, even when the  
corresponding end nodes are inactive.  
You can continue to add new static MAC addresses to a port operating  
under this security level.  
Note  
For background information on MAC addresses and aging time,  
352  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
AT-S63 Management Software Features Guide  
Invalid Frames and Intrusion Actions  
When a port receives an invalid frame, it has to select an intrusion action,  
which defines the port’s response to the packet. But before defining the  
intrusion actions, it helps to understand what constitutes an invalid frame.  
This differs for each security level, as explained here:  
ˆ Limited Security Level - An invalid frame for this security level is an  
ingress frame with a source MAC address not already learned by a  
port after the port had reached its maximum number of dynamic MAC  
addresses, or that was not assigned to the port as a static address.  
ˆ Secured Security Level - An invalid frame for this security level is an  
ingress frame with a source MAC address that was not entered as a  
static address on the port.  
ˆ Locked - An invalid frame for this security level is an ingress frame with  
a source MAC address that the port has not already learned or that  
was not assigned as a static address.  
Intrusion action defines what a port does when it receives an invalid frame.  
For a port operating under either the Secured or Locked security mode,  
the intrusion action is always the same. The port discards the frame.  
But with the Limited security mode you can specify the intrusion action.  
Here are the options:  
ˆ Discard the invalid frame.  
ˆ Discard the invalid frame and send an SNMP trap. (SNMP must be  
enabled on the switch for the trap to be sent.)  
ˆ Discard the invalid frame, send an SNMP trap, and disable the port.  
Section VIII: Port Security  
353  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 30: MAC Address-based Port Security  
Guidelines  
The following guidelines apply to MAC address-based port security:  
ˆ The filtering of a packet occurs on the ingress port, not on the egress  
port.  
ˆ You cannot use MAC address port security and 802.1x port-based  
access control on the same port. To configure a port as an  
Authenticator or Supplicant in 802.1x port-based access control, you  
must set its MAC address security level to Automatic, which is the  
default setting.  
ˆ This type of port security is not supported on optional GBIC, SFP, or  
XFP modules.  
ˆ All of a port’s static MAC addresses are deleted when its security level  
is changed from Locked to any of the other three security levels.  
354  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 31: 802.1x Port-based Network Access Control  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
356  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
The AT-S63 Management Software has several different methods for  
protecting your network and its resources from unauthorized access. For  
explains how you can restrict network access using the MAC addresses of  
the end nodes of your network.  
This chapter explains yet another way. This method, referred to as 802.1x  
port-based network access control, uses the RADIUS protocol to control  
who can send traffic through and receive traffic from a switch port. The  
switch does not allow an end node to send or receive traffic through a port  
until the user of the node has by authenticated by a RADIUS server.  
The benefit of this type of network security is obvious. You can use it to  
prevent unauthorized individuals from connecting a computer to a switch  
port or using an unattended workstation to access your network resources.  
Only those users designated as valid network users on the RADIUS server  
will be permitted to use the switch to access the network.  
This port security method uses the RADIUS authentication protocol. The  
AT-S63 Management Software is shipped with RADIUS client software. If  
page 423, then you know that you can use the RADIUS client software on  
the switch, along with a RADIUS server on your network, to also create  
new manager accounts that control who can manage and change the  
AT-S63 parameter on the switch.  
Note  
RADIUS with Extensible Authentication Protocol (EAP) extensions is  
the only supported authentication protocol for 802.1x Port-based  
Network Access Control. This feature is not supported with the  
TACACS+ authentication protocol. The switch supports only one  
authentication protocol at a time. Consequently, if you want to  
implement 802.1 Port-based Network Access Control and also  
create new manager accounts as explained in Chapter 36,  
RADIUS protocol.  
Following are several terms to keep in mind when you use this feature.  
ˆ Supplicant - A supplicant is an end user or end node that wants to  
access the network through a switch port. A supplicant is also referred  
to as a client.  
ˆ Authenticator - The authenticator is a port on the switch that prohibits  
network access by a supplicant until the supplicant has been validated  
by the RADIUS server.  
Section VIII: Port Security  
357  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 31: 802.1x Port-based Network Access Control  
ˆ Authentication server - The authentication server is the network device  
that has the RADIUS server software. This is the device that does the  
actual authenticating of the supplicants.  
The AT-9400 Switch does not authenticate any of the supplicants  
connected to its ports. It’s function is to act as an intermediary between a  
supplicant and the authentication server during the authentication  
process.  
358  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Authentication Process  
Below is a brief overview of the authentication process that occurs  
between a supplicant, authenticator, and authentication server. For further  
details, refer to the IEEE 802.1x standard.  
ˆ Either the authenticator (that is, a switch port) or the supplicant initiates  
an authentication message exchange. The switch initiates an  
exchange when it detects a change in the status of a port (such as  
when the port transitions from no link to valid link), or if it receives a  
packet on the port with a source MAC address not in the MAC address  
table.  
ˆ An authenticator starts the exchange by sending an EAP-Request/  
Identity packet. A supplicant starts the exchange with an EAPOL-Start  
packet, to which the authenticator responds with a EAP-Request/  
Identity packet.  
ˆ The supplicant responds with an EAP-Response/Identity packet to the  
authentication server via the authenticator.  
ˆ The authentication server responds with an EAP-Request packet to  
the supplicant via the authenticator.  
ˆ The supplicant responds with an EAP-Response/MD5 packet  
containing a username and password.  
ˆ The authentication server sends either an EAP-Success packet or  
EAP-Reject packet to the supplicant.  
ˆ Upon successful authorization of the supplicant by the authentication  
server, the switch adds the supplicant’s MAC address to the MAC  
address as an authorized address and begins forwarding network  
traffic to and from the port.  
ˆ When the supplicant sends an EAPOL-Logoff message, the switch  
removes the supplicant’s MAC address from the MAC address table,  
preventing the supplicant from sending or receiving any further traffic  
from the port.  
Section VIII: Port Security  
359  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 31: 802.1x Port-based Network Access Control  
Port Roles  
Part of the task of implementing this feature is specifying the roles of the  
ports on the switch. A port can have one of three roles:  
ˆ None  
ˆ Authenticator  
ˆ Supplicant  
None Role A switch port in the None role does not participate in port-based access  
control. Any device can connect to the port and send traffic through it and  
receive traffic from it without being validated. This port setting is  
appropriate if no validation is required for the network device connected to  
the port. This is the default setting for the switch ports.  
Note  
Because a RADIUS authentication server cannot authenticate itself,  
it must communicate with the switch through a port that is set to the  
None role.  
Authenticator Placing a switch port in the authenticator role activates port access control  
on the port. A port in the role of authenticator does not forward network  
traffic to or from the end node until the client has been authenticated by a  
RADIUS server.  
Role  
Determining whether a switch port should be set to the authenticator role  
is straightforward. You should set a port on a switch to the authenticator  
role if you want the user of the end node connected to the port to be  
authenticated before being permitted to use the network.  
Authentication Modes  
The AT-9400 Switch supports two authentication modes on an  
authenticator port.  
ˆ 802.1x username/password combination  
In this authentication mode, each supplicant connected to an  
authenticator port must be assigned a unique username and password  
combination on the RADIUS server. A supplicant must provide the  
information either manually or automatically when initially passing  
traffic through an authenticator port and during reauthentications. The  
802.1x client software on the supplicant either prompts the user for the  
necessary information or provides the information automatically.  
360  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
             
AT-S63 Management Software Features Guide  
Assigning unique username and password combinations to your  
network users and requiring the users to provide the information when  
they initially send traffic through the switch can enhance network  
security by limiting network access to only those supplicants who have  
been assigned valid combinations. Another advantage is that the  
authentication is not tied to any specific computer or node. An end user  
can log on from any system and still be verified by the RADIUS server  
as a valid user of the switch and network.  
This authentication method requires 802.1x client software on the  
supplicant nodes.  
ˆ MAC address-based authentication  
An alternative method is to use the MAC address of a node as the  
username and password combination for the device. The client is not  
prompted for this information. Rather, the switch extracts the source  
MAC address from the initial frames received from a supplicant and  
automatically sends the MAC address as both the username and  
password of the supplicant to the RADIUS server for authentication.  
The advantage to this approach is that the supplicant need not have  
802.1x client software. The disadvantage is that because the client is  
not prompted for a username and password combination, it does not  
guard against an unauthorized individual from gaining access to the  
network through an unattended network node or by counterfeiting a  
valid network MAC address.  
Operational Settings  
A port in the authenticator role can have one of three possible operational  
settings:  
ˆ Auto - Activates port-based authentication. The port begins in the  
unauthorized state, forwarding only EAPOL frames and discarding all  
other traffic. The authentication process begins when the link state of  
the port changes or the port receives an EAPOL-Start packet from a  
supplicant. The switch requests the identity of the client and begins  
relaying authentication messages between the client and the RADIUS  
authentication server. After the supplicant is validated by the RADIUS  
server, the port begins forwarding all traffic to and from the supplicant.  
This is the default setting for an authenticator port.  
ˆ Force-authorized - Disables IEEE 802.1X port-based authentication  
and automatically places the port in the authorized state without any  
authentication exchange required. The port transmits and receives  
normal traffic without authenticating the client.  
Section VIII: Port Security  
361  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 31: 802.1x Port-based Network Access Control  
Note  
A supplicant connected to an authenticator port set to force-  
authorized must have 802.1x client software if the port’s  
authenticator mode is 802.1x. Though the force-authorized setting  
prevents an authentication exchange, the supplicant must still have  
the client software to forward traffic through the port.  
ˆ Force-unauthorized - Causes the port to remain in the unauthorized  
state, ignoring all attempts by the supplicant to authenticate. The port  
forwards EAPOL frames, but discards all other traffic. This setting is  
analogous to disabling a port.  
As mentioned earlier, the switch itself does not authenticate the user  
names and passwords from the clients. That function is performed by the  
authentication server and the RADIUS server software. The switch acts as  
an intermediary for the authentication server by denying access to the  
network by the client until the client has been validated by the  
authentication server.  
Supplicant Role A switch port in the supplicant role acts as a client. The port assumes it  
must log in by providing a valid user name and password to whatever  
device it is connected to, typically another switch port.  
Figure 39 illustrates the port role. Port 11 on switch B has been set to the  
supplicant role. Now, whenever switch B is power cycled or reset and  
initiates a link with switch A, it must log on by providing a username and  
password. (You enter this information when you configure the port for the  
supplicant role.)  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
Switch A  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Port 6  
in  
Authenticator  
Role  
RADIUS  
Authentication  
Server  
Port 11  
in Supplicant Role  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
SFP  
23  
SFP  
24  
1000 LINK  
/
ACT  
Switch B  
L/A  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Figure 39. Example of the Supplicant Role  
362  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Authenticator Ports with Single and Multiple Supplicants  
An authenticator port has two operating modes. The modes relate to the  
number of clients using the port and, in situations where an authenticator  
port is supporting more than one client, whether just one client or all the  
clients must log on to use the switch port.  
The operating modes are:  
ˆ Single  
ˆ Multiple  
Single Operating The Single operating mode is used in two situations. The first is when an  
authenticator port supports only one client. In this scenario, the switch  
allows only one client to log on and use the port.  
Mode  
You can also use the Single mode when an authenticator port supports  
more than one client, but where only one client needs to log on in order for  
all clients to use the port. This configuration can be useful in situations  
where you want to add 802.1x Port-based Network Access Control to a  
switch port that is supporting multiple clients, but want to avoid having to  
create individual accounts for all the clients on the RADIUS server.  
This is referred to as “piggy-backing.” After one client has successfully  
logged, the port permits the other clients to piggy-back onto the initial  
client’s log on, allowing all clients to forward packets through the port.  
To implement this configuration, you have to set the operating mode of an  
authenticator port to Single and also toggle the piggy-back mode feature.  
When piggy-back is disabled, only one client is allowed to log on and use  
the port. When this feature is enabled, an unlimited number of clients can  
use the port after one client has successfully logged on.  
Note, however, that should the client who accomplished the initial log on  
fail to periodically reauthenticate or log out, the switch port reverts to the  
unauthenticated state. It bars all further traffic to and from all the clients on  
the port, until the initial client or another client logs on.  
Here are several examples that illustrate the Single operating mode and  
the piggy-back mode of an authenticator port. In Figure 40 on page 364,  
an authenticator port on a switch, in this case port 6, is connected to a  
single client. The authenticator port’s operating mode is set to Single and  
the piggy-back feature is disabled so that only one client can use the port  
at any one time.  
Section VIII: Port Security  
363  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 31: 802.1x Port-based Network Access Control  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
AT-9400 Switch  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
RADIUS  
Authentication  
Server  
Port 6  
Role: Authenticator  
Operating Mode: Single  
Piggy-back Mode: Disabled  
Authenticated Client  
Figure 40. Authenticator Port in Single Operating Mode with a Single  
Client  
the piggy-back mode. Multiple clients are connected to an authenticator  
port on the switch through an Ethernet hub or a non-802.1x-compliant  
Ethernet switch, such as an unmanaged switch. The operating mode of  
the authenticator port on the AT-9400 Switch is set to Single and the  
piggy-back mode is enabled. This allows all clients to forward packets  
through the port after one client logs on.  
364  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
AT-9400 Switch  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
RADIUS  
Authentication  
Server  
Port 6  
Role: Authenticator  
Operating Mode: Single  
Piggy-back Mode: Enabled  
Ethernet Hub or  
Non-802.1x-compliant  
Switch  
Unauthenticated  
Clients  
Authenticated  
Client  
Figure 41. Single Operating Mode with Multiple Clients Using the Piggy-  
back Feature - Example 1  
Because the piggy-back mode is activated on the authenticator port, only  
one client needs to be authenticated in order for all the clients to forward  
traffic through the port. If the port is using the 802.1x authentication  
method, then at least one client must have 802.1x client firmware and  
provide a username and password during authentication. (The other  
clients do not need 802.1x client firmware to forward traffic through the  
port after one client is authenticated.)  
If the switch port is set to MAC address-based authentication, 802.1 client  
firmware is not required. The MAC address of the first client to forward  
traffic through the port is used for authentication. When that client is  
authenticated, all supplicants have access to the port.  
As mentioned early, should the client who performed the initial log on fail  
to reauthenticate when necessary or log out, the port reverts to the  
unauthenticated state, blocking all traffic to and from all clients. Another  
client must be authenticated in order for all remaining clients to continue to  
forward traffic through the port.  
Section VIII: Port Security  
365  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 31: 802.1x Port-based Network Access Control  
If the clients are connected to an 802.1x-compliant device, such as  
another AT-9400 Switch, you can automate the initial log on and  
reauthentications by configuring one of the switch ports as a supplicant. In  
this manner, the log on and reauthentications are performed  
automatically, eliminating the need for relying on an individual to perform  
the task. This scenario is illustrated in Figure 42.  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
8
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
AT-9400 Switch (A)  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
23  
24  
POWER  
2
4
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Port 6  
Role: Authenticator  
Operating Mode: Single  
Piggy-back Mode: Enabled  
RADIUS  
Authentication  
Server  
Port 11  
Role: Supplicant  
Username: sales_switch  
Password: wind4411  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
AT-9400 Switch (B)  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Client Ports:  
Role: None  
Unauthenticated  
Clients  
Figure 42. Single Operating Mode with Multiple Clients Using the Piggy-  
back Feature - Example 2  
None of the workstations connected to switch B need to be authenticated  
or require 802.1x client software when accessing switch A because the log  
on to switch A and the subsequent reauthentications are performed  
automatically by the supplicant port on switch B, which is connected to an  
authenticator port on switch A with piggy-back mode enabled. It should be  
noted, however, that in this particular scenario the clients have full access  
to the resources of switch B even if the switch fails to log on or  
reauthenticate to switch A.  
The example in the next figure again illustrates two 802.1x-compliant  
switches. The primary difference between this and the previous example  
is that the clients in the previous example did not have to log on to access  
switch B. In this example the clients have to log on to have any access at  
all to the network.  
366  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
AT-9400 Switch (A)  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Port 6:  
Role: None  
or  
Role: Authenticator  
Operating Mode: Single  
Piggy-back Mode: Enabled  
RADIUS  
Authentication  
Server  
Port 11:  
Role: None  
or  
Role: Supplicant  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
ACT  
10/100 LINK  
HDX COL  
11 13 15 17 19 21 23R  
/
ACT  
LASER PRODUCT  
/
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
AT-9400 Switch (B)  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Client Ports:  
Role: Authenticator  
Operating Mode: Single  
Piggy-back Mode: Disabled  
Authenticated  
Clients  
Figure 43. Single Operating Mode with Multiple Clients Using the Piggy-  
back Feature - Example 3  
Multiple The second type of operating mode for an authenticator port is the Multiple  
mode. You can use this mode when a port is supporting more than one  
Operating Mode  
client and you want each client to log on individually before being  
permitted to forward traffic through the port. An authenticator port in this  
mode can support up to a maximum of 320 clients, with a total maximum  
of 480 per switch. If you are using the 802.1x authentication method, you  
must provide each client with a separate username and password  
combination and the clients must provide their combinations to forward  
traffic through a switch port.  
Selecting the Multiple mode for an authenticator port disables the piggy-  
back mode, because this operating mode does not permit piggy-backing.  
Section VIII: Port Security  
367  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 31: 802.1x Port-based Network Access Control  
An example of this authenticator operating mode is illustrated in Figure 44.  
The clients are connected to a hub or non-802.1x-compliant switch which  
is connected to an authenticator port on the AT-9400 Switch. If the  
authenticator port is set to the 802.1x authentication method, the clients  
must provide their username and password combinations before they can  
forward traffic through the AT-9400 Switch.  
If the authentication method is MAC address-based, the authenticator port  
uses the MAC addresses of the clients as the username and password  
combinations. The port accepts and forwards traffic only from those clients  
whose MAC addresses have been entered on the RADIUS server and  
denies access to all other users.  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
AT-9400 Switch  
23  
24  
POWER  
2
4
6
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
RADIUS  
Authentication  
Server  
Port 6  
Role: Authenticator  
Operating Mode: Multiple  
Piggy-back Mode: -----  
Ethernet Hub or  
Non-802.1x-compliant  
Switch  
Authenticated  
Clients  
Figure 44. Authenticator Port in Multiple Operating Mode - Example 1  
The next example of the multiple mode in Figure 45 shows two AT-9400  
Switches. The clients connected to switch B have to log on to port 6 on  
Switch A when they pass a packet to that switch for the first time.  
There are several items to note when interconnecting two 802.1x-  
compliant devices using the Multiple operating mode of an authenticator  
port. In order for switch B in our example to pass the RADIUS messages  
to switch A, it must be able to log on to port 6 on switch A. That is why port  
11 on the lower switch is configured as a supplicant. If its role is set to  
368  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
none, port 6 on switch A will discard the packets because switch B would  
not be logged on to the port.  
Also notice that the ports where the clients are connected on switch B are  
set to the none role. This is because a client can log on only once. If, in this  
example, you were to make a client’s port an authenticator, the client  
would have to log on twice when trying to access switch A, once on its port  
on switch B as well as the authenticator port on switch A. This is not  
permitted. Consequently, in our example the clients on switch B have full  
access to that switch, but are denied access to switch A until they log on to  
port 6 on switch A.  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
8
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
AT-9400 Switch (A)  
23  
24  
POWER  
2
4
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Port 6  
Role: Authenticator  
Operating Mode: Multiple  
Piggy-back Mode: -----  
RADIUS  
Authentication  
Server  
Port 11  
Role: Supplicant  
Username: switch24  
Password: waveform  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
CLASS  
1
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
AT-9400 Switch (B)  
23  
24  
POWER  
2
4
8
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Client Ports:  
Role: None  
Authenticated  
Clients  
Figure 45. Authenticator Port in Multiple Operating Mode - Example 2  
Section VIII: Port Security  
369  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 31: 802.1x Port-based Network Access Control  
Supplicant and VLAN Associations  
One of the challenges to managing a network is accommodating end  
users that roam. These are individuals whose work requires that they  
access the network resources from different points at different times. The  
difficulty arises in providing them with access to the same network  
resources and, conversely, restricting them from unauthorized areas,  
regardless of the workstation from where they access the network. A  
closely related issue is where a workstation is employed at various times  
by different individuals with unique requirements in terms of network  
resources and security levels.  
Providing network users with access to their network resources while also  
maintaining network security is often achieved through the use of VLANs.  
As explained in “Overview” on page 249, a VLAN is an independent traffic  
domain where the traffic generated by the nodes within the VLAN is  
restricted to nodes of the same VLAN, unless there is a router or Layer 3  
device. Different users are assigned to different VLANs depending on their  
resource requirements and security level.  
The problem with a port-based VLAN is that VLAN membership is  
determined by the port on the switch to which the device is connected. If a  
different device that needs to belong to a different VLAN is connected to  
the port, the port must be manually moved to the new VLAN using the  
management software.  
With 802.1x port-based network access control, you can link a username  
and password combination or MAC address to a specific VLAN so that the  
switch automatically moves the port to the appropriate VLAN when a client  
logs on. This frees the network manager from having to reconfigure  
VLANs as end users access the network from different points or where the  
same workstation is used by different individuals at different times.  
To use this feature, you have to enter a VLAN identifier, along with other  
information, when you create a supplicant account on the RADIUS server.  
The server passes the identifier to the switch when a user logs on with a  
valid username and password combination or MAC address, depending  
on the authentication method. The information to provide on the RADIUS  
How the switch responses when it receives VLAN information during the  
authentication process can differ depending on the operating mode of the  
authenticator port.  
370  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Single Operating Here are the operating characteristics for the switch when an authenticator  
port is set to the Single operating mode:  
Mode  
ˆ If the switch receives a valid VLAN ID or VLAN name from the RADIUS  
server, it moves the authenticator port to the designated VLAN and  
changes the port to the authorized state. If the piggy-back mode is  
disabled, only the authenticated supplicant is allowed to use the port.  
All other supplicants are denied entry. If the piggy-back mode is  
enabled, all clients are allowed access to the port and the same VLAN  
after the initial authentication.  
ˆ If the switch receives an invalid VLAN ID or VLAN name from the  
RADIUS server (e.g., the VID of a nonexistent VLAN), it leaves the port  
in the unauthorized state to deny access to the port.  
Multiple The initial authentication on an authenticator port running in the Multiple  
operating mode is handled in the same fashion as with the Single  
Operating Mode  
operating mode. If the switch receives a valid VLAN ID or name from the  
RADIUS server, it moves the authenticator port to the designated VLAN  
and changes the port to the authorized state.  
How the switch handles subsequent authentications on the same port  
depends on how you set the Secure VLAN parameter. Your options are as  
follows:  
ˆ If you activate the Secure VLAN feature, only those supplicants with  
the same VLAN assignment as the initial supplicant are authenticated.  
Supplicants with a different VLAN assignment or with no VLAN  
assignment are denied access to the port.  
ˆ If you disable the Secure VLAN feature, all supplicants, regardless of  
their assigned VLANs, are authenticated. However, the port remains in  
the VLAN specified in the initial authentication.  
Supplicant VLAN The following information must be entered as part of a supplicant’s  
account on the RADIUS server when associating a supplicant to a VLAN.  
Attributes on the  
RADIUS Server  
ˆ Tunnel-Type  
The protocol to be used by the tunnel specified by Tunnel-Private-  
Group-Id. The only supported value is VLAN (13).  
ˆ Tunnel-Medium-Type  
The transport medium to be used for the tunnel specified by Tunnel-  
Private-Group-Id. The only supported value is 802 (6).  
ˆ Tunnel-Private-Group-ID  
The ID of the tunnel the authenticated user should use. This must be  
the name of VID of the VLAN of the switch.  
Section VIII: Port Security  
371  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 31: 802.1x Port-based Network Access Control  
Guest VLAN  
An authenticator port in the unauthorized state typically accepts and  
transmits only 802.1x packets while waiting to authenticate a supplicant.  
However, you can configure an authenticator port to be a member of a  
Guest VLAN when no supplicant is logged on. Any client using the port is  
not required to log on and has full access to the resources of the Guest  
VLAN.  
If the switch receives 802.1x packets on the port, signalling that a  
supplicant is logging on, it moves the port to its predefined VLAN and  
places it in the unauthorized state. The port remains in the unauthorized  
state until the log on process between the supplicant and the RADIUS  
server is completed. When the supplicant logs off, the port automatically  
returns to the Guest VLAN.  
Note  
The Guest VLAN feature is only supported on an authenticator port  
in the Single operating mode.  
372  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
RADIUS Accounting  
The AT-S63 Management Software supports RADIUS accounting for  
switch ports set to the Authenticator role. This feature sends information to  
the RADIUS server about the status of its supplicants. You can view this  
information on the RADIUS server to monitor network activity and use.  
The switch sends accounting information to the RADIUS server when one  
of the following events occur:  
ˆ Supplicant logs on  
ˆ Supplicant logs off  
ˆ A change in the status of an Authenticator port during an active  
Supplicant session (for example, the port is reset or is changed from  
the Authenticator role to None role while a Supplicant is logged on)  
The information sent by the switch to the RADIUS server for an event  
includes:  
ˆ Port number where the event occurred  
ˆ The date and time when the event occurred  
ˆ The number of packets transmitted and received by the switch port  
during a supplicant’s session. (This information is sent only when the  
client logs off.)  
You can also configure the accounting feature to send interim updates so  
you can monitor which clients are still active.  
Here are the guidelines to using the accounting feature:  
ˆ The AT-S63 Management Software supports the Network level of  
accounting, but not the System or Exec.  
ˆ This feature is only available for ports operating in the Authenticator  
role. No accounting is provided for ports operating in the Supplicant or  
None role.  
ˆ You must configure 802.1x Port-based Network Access Control as  
explained in this chapter and designate the Authenticator ports.  
ˆ You must also specify from one to three RADIUS servers.  
Section VIII: Port Security  
373  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 31: 802.1x Port-based Network Access Control  
General Steps  
Here are the general steps to implementing 802.1x Port-based Network  
Access Control and RADIUS accounting on the switch:  
1. You must install a RADIUS server on one or more of your network  
servers or management stations. Authentication protocol server  
software is not available from Allied Telesis. Funk Software Steel-  
Belted Radius and Free Radius have been verified as fully compatible  
with the AT-S63 Management Software.  
Note  
This feature is not supported with the TACACS+ authentication  
protocol.  
2. Those clients connected to an authenticator port set to the 802.1x  
authentication method will need 802.1x client software. Microsoft  
WinXP client software and Meeting House Aegis client software have  
been verified as fully compatible with the AT-S63 Management  
Software. (802.1x client software is not required when an authenticator  
port is set to the MAC address-based authentication method.)  
3. You must configure and activate the RADIUS client software in the  
AT-S63 Management Software. The default setting for the  
authentication protocol is disabled. You will need to provide the  
following information:  
ˆ The IP addresses of up to three RADIUS servers.  
ˆ The encryption key used by the authentication servers.  
4. You must configure the port access control settings on the switch. This  
involves the following:  
ˆ Specifying the port roles.  
ˆ Configuring 802.1x port parameters.  
ˆ Enabling 802.1x Port-based Network Access Control.  
The instructions for this step are found in this chapter.  
5. If you want to use RADIUS accounting to monitor the clients  
connected to the switch ports, you must configure the service on the  
switch.  
374  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Guidelines  
The following are general guidelines to using this feature:  
ˆ Ports operating under port-based access control do not support  
dynamic MAC address learning.  
ˆ The appropriate port role for a port on the AT-9400 Switch connected  
to a RADIUS authentication server is None.  
ˆ The authentication method of an authenticator port can be either  
802.1x username and password combination or MAC address-based,  
but not both.  
ˆ A supplicant must have 802.1x client software if the authentication  
method of a switch port is 802.1x username and password  
combination.  
ˆ A supplicant does not need 802.1x client software if the authentication  
method of an authenticator port is MAC address-based.  
ˆ An authenticator port set to the multiple operating mode can support  
up to a maximum of 320 authenticated supplicants at one time.  
ˆ The switch can handle up to a maximum of 480 authenticated  
supplicants at one time. The switch stops accepting new  
authentications after the maximum is reached and starts accepting  
new authentications as supplicants log out or are timed out.  
ˆ An 802.1x username and password combination is not tied to the MAC  
address of an end node. This allows end users to use the same  
username and password when working at different workstations.  
ˆ After a client has successfully logged on, the MAC address of the end  
node is added to the switch’s MAC address table as an authenticated  
address. It remains in the table until the client logs off the network or  
fails to reauthenticate, at which point the address is removed. The  
address is not timed out, even if the node becomes inactive.  
Note  
End users of 802.1x port-based network access control should be  
instructed to always log off when they are finished with a work  
session. This can prevent unauthorized individuals from accessing  
the network through unattended network workstations.  
ˆ Authenticator and supplicant ports must be untagged ports. They  
cannot be tagged ports of any VLAN.  
ˆ The MAC address-based port security setting for an authenticator port  
must be Automatic. This restriction does not apply to a supplicant port.  
For further information, refer to Chapter 30, “MAC Address-based Port  
Section VIII: Port Security  
375  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 31: 802.1x Port-based Network Access Control  
ˆ An authenticator port cannot be part of a static port trunk, LACP port  
trunk, or port mirror.  
ˆ If a switch port set to the supplicant role is connected to a port on  
another switch that is not set to the authenticator role, the port, after a  
timeout period, assumes that it can send traffic without having to log  
on.  
ˆ GVRP must be disabled on an authenticator port.  
ˆ When 802.1x port-based network access control is activated on a  
switch, the feature polls all RADIUS servers specified in the RADIUS  
configuration. If three servers have been configured, the switch polls  
all three. If server 1 responds, all future requests go only to that server.  
If server 1 stops responding, the switch again polls all RADIUS  
servers. If server 2 responds, but not server 1, then all future requests  
go to servers 1 and 2. If only server 3 responds, then all future  
requests go to all three servers.  
ˆ In order to change the untagged VLAN assignment of an authenticator  
or supplicant port, you must set its port role to none. You can change  
the port’s role back to authenticator or supplicant after you have  
changed the port’s VLAN assignment.  
ˆ To use the Guest VLAN feature, the designated VLAN must already  
exist on the switch.  
ˆ A Guest VLAN can be either port-based or tagged.  
ˆ The switch must be running in the user-configured VLAN mode to  
support 802.1x port-based network access control. The feature is not  
supported in the multiple VLAN modes.  
ˆ Starting with version 3.0.0, the AT-S63 Management Software  
supports EAP-MD5, EAP-TLS, EAP-TTLS, and EAP-PEAP  
authentication for both authenticators and supplicants. Prior to version  
3.0.0, only EAP-MD5.was supported.  
ˆ The switch must have a routing interface on the local subnet where the  
RADIUS server is a member. The switch uses the routing interface’s IP  
address as its source address when communicating with the server.  
For background information, refer to “Routing Interfaces and  
Note  
Prior to version 2.0.0 of the AT-S63 Management Software, the  
RADIUS server had to be a member of the switch’s management  
VLAN. This restriction no longer applies. The server can be located  
on any local subnet on the switch that has a routing interface.  
376  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Here are guidelines for adding VLAN assignments to supplicant accounts  
on a RADIUS server:  
ˆ The VLAN can be either port-based or tagged.  
ˆ The VLAN must already exist on the switch.  
ˆ A client can have only one VLAN associated with it on the RADIUS  
server.  
ˆ When a supplicant logs on, the switch port is moved as an untagged  
port to the designated VLAN.  
Section VIII: Port Security  
377  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 31: 802.1x Port-based Network Access Control  
378  
Section VIII: Port Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Section IX  
Management Security  
The chapters in this section describe the management security features of  
the AT-9400 Switch. The chapters include:  
Section IX: Management Security  
379  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
380  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 32  
Web Server  
The sections in this chapter are:  
Section IX: Management Security  
381  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 32: Web Server  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from the following management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
382  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Overview  
The AT-S63 Management Software has a web server and a special web  
browser interface that provide the ability to remotely manage the switch  
from a management workstation on your network using a web browser.  
(For instructions on the switch’s web browser interface, refer to the  
AT-S63 Management Software Web Browser Interface User’s Guide.)  
The web server on the switch can operate in HTTP or HTTPS mode. A  
management session conducted in the HTTP mode is not secure because  
the packets are transmitted in plaintext, including the manager’s login  
name and password. Should someone be monitoring the traffic on your  
network during a management session, the security of the unit may be  
jeopardize.  
In contrast, a management session conducted in the HTTPS mode is  
secure because the load in the management packets is encrypted with the  
Secure Sockets Layer (SSL) protocol. This mode requires an encryption  
key pair and a certificate. For background information, refer to Chapter 33,  
The default setting for the web server is disabled, with the non-secure  
HTTP mode as the default active mode.  
For background information and guidelines on remote management, refer  
to the Starting an AT-S63 Management Session Guide.  
Note  
To use HTTPS in an enhanced stack, all switches in the stack must  
use HTTPS. For further information, refer to “SSL and Enhanced  
Supported The switch supports the following HTTP and HTTPs protocols:  
Protocols  
ˆ HTTP v1.0 and v1.1 protocols  
ˆ HTTPS v1.0 and v1.1 protocols running over SSL  
The switch supports the following SSL protocols:  
ˆ SSL version 2.0  
ˆ SSL version 3.0  
ˆ TLS (Transmission Layer Security) version 1.0  
Section IX: Management Security  
383  
Download from Www.Somanuals.com. All Manuals Search And Download.  
         
Chapter 32: Web Server  
Configuring the Web Server for HTTP  
The following steps configure the web server for non-secure HTTP  
operation. The steps reference only the command line commands, but the  
web server can be configured from the menus interface, too.  
1. Disable the web server with the DISABLE HTTP SERVER command.  
2. Activate HTTP in the web server with the SET HTTP SERVER  
command.  
3. Enable the web server with the ENABLE HTTP SERVER command.  
384  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Configuring the Web Server for HTTPS  
The following sections outline the steps for configuring the web server on  
the switch for HTTPS operation with a self-signed or CA certificate. The  
steps reference only the command line commands, but the web server can  
be configured from the menus interface, too.  
General Steps for These steps configure the web server with a self-signed certificate:  
a Self-signed  
1. Set the switch’s date and time. The date and time are stamped in the  
Certificate  
certificate.  
2. Create a public and private key pair with the CREATE ENCO KEY  
command.  
3. Create a self-signed certificate using the public and private key pair  
with the CREATE PKI CERTIFICATE command.  
4. Add the certificate to the certificate database with the ADD PKI  
CERTIFICATE command.  
5. Disable the web server with the DISABLE HTTP SERVER command.  
6. Activate HTTPS in the web server with the SET HTTP SERVER  
command.  
7. Enable the web server with the ENABLE HTTP SERVER command.  
For an example of this command sequence, refer to the SET HTTP  
SERVER command in the AT-S63 Management Software Command Line  
Interface User’s Guide.  
General Steps for These steps configure the web server with a public or private CA  
certificate.  
a Public or  
Private CA  
Certificate  
1. Set the switch’s date and time. The date and time are stamped in the  
enrollment request.  
2. Create a public and private key pair with the CREATE ENCO KEY  
command.  
3. Generate an enrollment request with the CREATE PKI  
ENROLLMENTREQUEST command.  
4. Upload the enrollment request from the switch’s file system with the  
UPLOAD METHOD=TFTP or UPLOAD METHOD=XMODEM  
command.  
5. Submit the enrollment request to a public or private CA.  
Section IX: Management Security  
385  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 32: Web Server  
6. After receiving the certificates from the CA, download them into the  
switch’s file system using the LOAD METHOD=TFTP or LOAD  
METHOD=XMODEM command.  
7. Add the certificates to the certificate database with the ADD PKI  
CERTIFICATE command.  
8. Disable the web server with the DISABLE HTTP SERVER command.  
9. Activate HTTPS in the web server with the SET HTTP SERVER  
command.  
10. Enable the web server with the ENABLE HTTP SERVER command.  
For an example of this command sequence, refer to the SET HTTP  
SERVER command in the AT-S63 Management Software Command Line  
Interface User’s Guide.  
386  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 33  
Encryption Keys  
The sections in this chapter are:  
For an overview of the procedures to configuring the switch’s web server  
Section IX: Management Security  
387  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 33: Encryption Keys  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from the following management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
You can view but not create encryption keys from the web browser  
interface.  
388  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
Protecting your managed switches from unauthorized management  
access is an important role for a network manager. Network operations  
and security can be severely compromised if an intruder gains access to  
critical switch information, such as a manager’s login username and  
password, and uses that information to alter a switch’s configuration  
settings.  
One way an intruder could obtain critical switch information is by  
monitoring network traffic with a network analyzer, such as a sniffer, and  
capturing management packets from remote Telnet or web browser  
management sessions. The payload in the packets exchanged during  
remote management sessions is transmitted in plaintext. The information  
obtained from the management packets could enable an intruder to  
access a switch.  
A way to prevent this type of assault is by encrypting the payload in the  
packets exchanged during a remote management session between a  
management station and a switch. Encryption makes the packets  
unintelligible to an outside agent. Only the remote workstation and the  
switch engaged in the management session are able to decode each  
other’s packets.  
A fundamental part of encryption is the encryption key. The key converts  
plaintext into encrypted text, and back again. A key consists of two  
separate keys: a private key and a public key. Together they create a key  
pair.  
The AT-S63 Management Software supports encryption for remote web  
browser management sessions using the Secure Sockets Layer (SSL)  
protocol. Adding encryption to your web browser management sessions  
involves creating one key pair and adding the public key of the key pair to  
a certificate, a digital document stored on the switch. You can have the  
switch create the certificate itself or you can have a public or private  
certificate authority (CA) create it for you. For an overview of the steps for  
adding encryption to your web browser management sessions, refer to  
The Telnet protocol does not support encryption. To employ encryption  
when remotely managing a switch using the menus interface, you must  
first obtain a Secure Shell (SSH) protocol application. SSH offers the same  
function as Telnet, but with encryption.  
SSH encryption requires that you create two key pairs on the switch— a  
server key pair and a host key pair and then configure the Secure Shell  
protocol server software on the switch, as explained in Chapter 35,  
Section IX: Management Security  
389  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 33: Encryption Keys  
Encryption Key Length  
When you create a key pair, you have to specify its length in bits. The  
range is 512, the default, to 1,536 bits, in increments of 256 bits. The  
longer the key, the more difficult it is for someone to decipher. If you are  
particularly concerned about the safety of your management sessions,  
you might want to use a longer key length than the default, though the  
default is likely to be sufficient in most situations.  
Creating a key is a very CPU intensive operation for the switch. Although  
the switch does not stop forwarding packets between the ports, the  
process can impact the CPU’s handling of network events, such as the  
processing of spanning tree BPDU packets, which can result in  
unexpected and unwanted switch behavior.  
A key with the default length should take the switch less than a minute to  
create. Longer keys can take up to 15 minutes. You should take this into  
account when creating a key to minimize the impact to the operations of  
your network. If you intend to create a long key, consider creating it before  
you connect the switch to the network, or during periods of low network  
traffic.  
390  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Encryption Key Guidelines  
Observe the following guidelines when creating an encryption key pair:  
ˆ Web browser encryption requires only one key pair.  
ˆ SSH encryption requires two key pairs. The keys must be of different  
lengths of at least one increment (256 bits) apart. The recommended  
size for the server key is 768 bits and the recommended size for the  
host key is 1024 bits.  
ˆ The AT-9400 Switch can only use those key pairs it has generated  
itself. The switch cannot use a key created on another system and  
imported onto the switch.  
ˆ The AT-S63 Management Software does not allow you to copy or  
export a private key from a switch. However, you can export a public  
key.  
ˆ The AT-S63 Management Software uses the RSA public key  
algorithm.  
ˆ Web browser and SSH encryption can share a key pair on the switch.  
Section IX: Management Security  
391  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 33: Encryption Keys  
Technical Overview  
The encryption feature provides the following data security services:  
ˆ Data encryption  
ˆ Data authentication  
ˆ Key exchange algorithms  
ˆ Key creation and storage  
Data Encryption Data encryption for switches is driven by the need for organizations to  
keep sensitive data private and secure. Data encryption operates by  
applying an encryption algorithm and key to the original data (the  
plaintext) to convert it into an encrypted form (the ciphertext). The  
ciphertext produced by encryption is a function of the algorithm used and  
the key. Because it is easy to discover what type of algorithm is being  
used, the security of an encryption system relies on the secrecy of its key  
information. When the ciphertext is received by the remote router, the  
decryption algorithm and key are used to recover the original plaintext.  
Often, a checksum is added to the data before encryption. The checksum  
allows the validity of the data to be checked on decryption.  
There are two main classes of encryption algorithm in use: symmetrical  
encryption and asymmetrical encryption.  
Symmetrical Encryption  
Symmetrical encryption refers to algorithms in which a single key is used  
for both the encryption and decryption processes. Anyone who has access  
to the key used to encrypt the plaintext can decrypt the ciphertext.  
Because the encryption key must be kept secret to protect the data, these  
algorithms are also called private, or secret key algorithms. The key can  
be any value of the appropriate length.  
DES Encryption Algorithms  
The most common symmetrical encryption system is the Data Encryption  
Standard (DES) algorithm (FIPS PUB 46). The DES algorithm has  
withstood the test of time and proved itself to be a highly secure  
encryption algorithm. To fully conform to the DES standard, the actual  
data encryption operations must be carried out in hardware. Software  
implementations can only be DES-compatible, not DES-compliant. The  
DES algorithm has a key length of 56 bits and operates on 64-bit blocks of  
data. DES can be used in the following modes:  
ˆ Electronic Code Book (ECB) is the fundamental DES function.  
Plaintext is divided into 64-bit blocks which are encrypted with the DES  
392  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
algorithm and key. For a given input block of plaintext ECB always  
produces the same block of ciphertext.  
ˆ Cipher Block Chaining (CBC) is the most popular form of DES  
encryption. CBC also operates on 64-bit blocks of data, but includes a  
feedback step which chains consecutive blocks so that repetitive  
plaintext data, such as ASCII blanks, does not yield identical  
ciphertext. CBC also introduces a dependency between data blocks  
which protects against fraudulent data insertion and replay attacks.  
The feedback for the first block of data is provided by a 64-bit  
Initialization Vector (IV). This is the DES mode used for the switch’s  
data encryption process.  
ˆ Cipher FeedBack (CFB) is an additive-stream-cipher method which  
uses DES to generate a pseudo-random binary stream that is  
combined with the plaintext to produce the ciphertext. The ciphertext is  
then fed back to form a portion of the next DES input block.  
ˆ Output FeedBack (OFB) combines the first IV DES algorithms with  
the plaintext to form ciphertext. The ciphertext is then used as the next  
IV.  
The DES algorithm has been optimized to produce very high speed  
hardware implementations, making it ideal for networks where high  
throughput and low latency are essential.  
Triple DES Encryption Algorithms  
The Triple DES (3DES) encryption algorithm is a simple variant on the  
DES CBC algorithm. The DES function is replaced by three rounds of that  
function, an encryption followed by a decryption followed by an encryption.  
This can be done by using either two DES keys (112-bit key) or three DES  
keys (168-bit key).  
The two-key algorithm encrypts the data with the first key, decrypts it with  
the second key and then encrypts the data again with the first key. The  
three-key algorithm uses a different key for each step. The three-key  
algorithm is the most secure algorithm due to the long key length.  
There are several modes in which Triple DES encryption can be  
performed. The two most common modes are:  
ˆ Inner CBC mode encrypts the entire packet in CBC mode three times  
and requires three different initial is at ion vectors (IV’s).  
ˆ Outer CBC mode triple encrypts each 8-byte block of a packet in CBC  
mode three times and requires one IV.  
Asymmetrical (Public Key) Encryption  
Asymmetrical encryption algorithms use two keys—one for encryption and  
one for decryption. The encryption key is called the public key because it  
cannot be used to decrypt a message and therefore does not need be kept  
Section IX: Management Security  
393  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 33: Encryption Keys  
secret. Only the decryption, or private key, needs to be kept secret. The  
other name for this type of algorithm is public key encryption. The public  
and private key pair cannot be randomly assigned, but must be generated  
together. In a typical scenario, a decryption station generates a key pair  
and then distributes the public key to encrypting stations. This distribution  
does not need to be kept secret, but it must be protected against the  
substitution of the public key by a malicious third party. Another use for  
asymmetrical encryption is as a digital signature. The signature station  
publishes its public key, and then signs its messages by encrypting them  
with its private key. To verify the source of a message, the receiver  
decrypts the messages with the published public key. If the message that  
results is valid, then the signing station is authenticated as the source of  
the message.  
The most common asymmetrical encryption algorithm is RSA. This  
algorithm uses mathematical operations which are relatively easy to  
calculate in one direction, but which have no known reverse solution. The  
security of RSA relies on the difficulty of factoring the modulus of the RSA  
key. Because key lengths of 512 bits or greater are used in public key  
encryption systems, decrypting RSA encrypted messages is almost  
impossible using current technology. The AT-S63 Management Software  
uses the RSA algorithm.  
Asymmetrical encryption algorithms require enormous computational  
resources, making them very slow when compared to symmetrical  
algorithms. For this reason they are normally only used on small blocks of  
data (for example, exchanging symmetrical algorithm keys), and not for  
entire data streams.  
Data Data authentication for switches is driven by the need for organizations to  
verify that sensitive data has not been altered.  
Authentication  
Data authentication operates by calculating a message authentication  
code (MAC), commonly referred to as a hash, of the original data and  
appending it to the message. The MAC produced is a function of the  
algorithm used and the key. Because it is easy to discover what type of  
algorithm is being used, the security of an authentication system relies on  
the secrecy of its key information. When the message is received by the  
remote switch, another MAC is calculated and checked against the MAC  
appended to the message. If the two MACs are identical, the message is  
authentic.  
Typically a MAC is calculated using a keyed one-way hash algorithm. A  
keyed one-way hash function operates on an arbitrary-length message  
and a key. It returns a fixed length hash. The properties which make the  
hash function one-way are:  
ˆ It is easy to calculate the hash from the message and the key  
ˆ It is very hard to compute the message and the key from the hash  
394  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
ˆ It is very hard to find another message and key which give the same  
hash  
The two most commonly used one-way hash algorithms are MD5  
(Message Digest 5, defined in RFC 1321) and SHA-1 (Secure Hash  
Algorithm, defined in FIPS-180-1). MD5 returns a 128-bit hash and SHA-1  
returns a 160-bit hash. MD5 is faster in software than SHA-1, but SHA-1 is  
generally regarded to be slightly more secure.  
HMAC is a mechanism for calculating a keyed Message Authentication  
Code which can use any one-way hash function. It allows for keys to be  
handled the same way for all hash functions and it allows for different  
sized hashes to be returned.  
Another method of calculating a MAC is to use a symmetric block cipher  
such as DES in CBC mode. This is done by encrypting the message and  
using the last encrypted block as the MAC and appending this to the  
original message (plain-text). Using CBC mode ensures that the whole  
message affects the resulting MAC.  
Key Exchange Key exchange algorithms are used by switches to securely generate and  
exchange encryption and authentication keys with other switches. Without  
Algorithms  
key exchange algorithms, encryption and authentication session keys  
must be manually changed by the system administrator. Often, it is not  
practical to change the session keys manually. Key exchange algorithms  
enable switches to re-generate session keys automatically and on a  
frequent basis.  
The most important property of any key exchange algorithm is that only  
the negotiating parties are able to decode, or generate, the shared secret.  
Because of this requirement, public key cryptography plays an important  
role in key exchange algorithms. Public key cryptography provides a  
method of encrypting a message which can only be decrypted by one  
party. A switch can generate a session key, encrypt the key using public  
key cryptography, transmit the key over an insecure channel, and be  
certain that the key can only be decrypted by the intended recipient.  
Symmetrical encryption algorithms can also be used for key exchange, but  
commonly require an initial shared secret to be manually entered into all  
switches in the secure network.  
The Diffie-Hellman algorithm, which is used by the AT-S63 Management  
Software, is one of the more commonly used key exchange algorithms. It  
is not an encryption algorithm because messages cannot be encrypted  
using Diffie-Hellman. Instead, it provides a method for two parties to  
generate the same shared secret with the knowledge that no other party  
can generate that same value. It uses public key cryptography and is  
commonly known as the first public key algorithm. Its security is based on  
the difficulty of solving the discrete logarithm problem, which can be  
compared to the difficulty of factoring very large integers.  
Section IX: Management Security  
395  
Download from Www.Somanuals.com. All Manuals Search And Download.  
         
Chapter 33: Encryption Keys  
A Diffie-Hellman algorithm requires more processing overhead than RSA-  
based key exchange schemes, but it does not need the initial exchange of  
public keys. Instead, it uses published and well tested public key values.  
The security of the Diffie-Hellman algorithm depends on these values.  
Public key values less than 768 bits in length are considered to be  
insecure.  
A Diffie-Hellman exchange starts with both parties generating a large  
random number. These values are kept secret, while the result of a public  
key operation on the random number is transmitted to the other party. A  
second public key operation, this time using the random number and the  
exchanged value, results in the shared secret. As long as no other party  
knows either of the random values, the secret is safe.  
396  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 34  
PKI Certificates and SSL  
The sections in this chapter are:  
Section IX: Management Security  
397  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 34: PKI Certificates and SSL  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from the following management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
You can view the PKI certificates and the SSL and PKI settings from the  
web browser interface, but you can not create or delete certificates; create  
or delete certificate enrollment requests; or change the settings from that  
interface.  
398  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
This chapter describes the second part of the encryption feature of the  
AT-S63 Management Software—PKI certificates. The first part is  
and certificates allow you to encrypt the communications between your  
management station and a switch during a web browser management  
session, and so protect your switch from intruders who might be using a  
sniffer to monitor the network for management packets.  
Types of Certificates  
As explained in the previous chapter, an encryption key encrypts the  
information in the frames exchanged between a switch and a web browser  
during a web browser management session. An encryption key consists of  
two parts: a private key and a public key. The private key remains on the  
switch and is used by the device to encrypt its messages.  
The public key is incorporated into a certificate and is used by your  
management station when you perform a web browser management  
session. Your web browser downloads the certificate with the public key  
from the switch when you begin a management session.  
The quickest and easiest way to create a certificate is to have the switch  
create it. This type of certificate is called a self-signed certificate. If you  
have a small to medium sized network, this will probably be the best  
approach. To review all the steps to configuring the web server for a self-  
Another option is to create the key but have someone else issue the  
certificate. That person, group, or organization is called a certification  
authority (CA).  
There are two kinds of CAs: public and private. A public CA issues  
certificates typically intended for use by the general public for other  
companies and organizations. A public CA requires proof of the identify of  
the company or organization before issuing a certificate. VeriSign is an  
example of a public CA.  
Because a certificate for the AT-9400 Switch is not intended for general  
use and will only be used by you and other network managers in managing  
the switch, it probably will not be necessary for you to have a public CA  
issue the certificate for the switch.  
Some large companies have private CAs. This is a person or group within  
the company that is responsible for issuing certificates for the company’s  
Section IX: Management Security  
399  
Download from Www.Somanuals.com. All Manuals Search And Download.  
           
Chapter 34: PKI Certificates and SSL  
network equipment. With private CAs, companies can keep track of the  
certificates and control access to various network devices.  
If your company is large enough, it might have a private CA and you might  
want the group to issue the certificate for the AT-9400 Switch so that you  
are in compliance with company policy.  
The first step to creating a CA certificate is to create a key pair. After that  
you must generate a digital document called an enrollment request and  
send the document to the CA. The document contains the public key and  
other information that the CA will use to create the certificate.  
Before sending an enrollment request to a CA, it is best to first contact the  
CA to determine what other documents or procedures might be required in  
order for the CA to create the certificate. This is particularly important with  
public CAs, which typically have strict guidelines on issuing certificates.  
400  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Distinguished Names  
Part of the task to creating a self-signed certificate or enrollment request is  
selecting a distinguished name. A distinguished name is integrated into a  
certificate along with the key and can have up to five parts. The parts are:  
ˆ cn - common name  
This can be the name of the person who will use the certificate.  
ˆ ou - organizational unit  
This is the name of a department, such as Network Support or IT.  
ˆ o - organization  
This is the name of the company.  
ˆ st - state  
This is the state.  
ˆ c - country  
This is the country  
A certificate name does not need to contain all of these parts. You can use  
as many or as few as you want. You separate the parts with a comma. You  
can use alphanumeric characters, as well as spaces in the name strings.  
You cannot use quotation marks. To use the following special characters  
{=,+<>#;\<CR>}, type a “\” before the character.  
Following are a few examples. This distinguished name contains only one  
part, the name of the switch:  
cn=Production Switch  
This distinguished name omits the common name, but includes everything  
else:  
ou=Network Support,o=XYZ Inc.,st=CA,c=US  
So what would be a good distinguished name for a certificate for the  
AT-9400 Switch? If the switch has an IP address, such as a master switch  
of an enhanced stack, you could use its address as the name. The  
following example is a distinguished name for a certificate for a master  
switch with the IP address 149.11.11.11:  
cn=149.11.11.11  
Section IX: Management Security  
401  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 34: PKI Certificates and SSL  
If your network has a Domain Name System and you mapped a name to  
the IP address of a switch, you can specify the switch’s name instead of  
the IP address as the distinguished name.  
For those switches that do not have an IP address, such as slave switches  
of an enhanced stack, you could assign their certificates a distinguished  
name using the IP address of the master switch of the enhanced stack.  
There is a benefit to giving a certificate a distinguished name equivalent to  
a switch’s IP address or domain name. This relates to what happens when  
you start a web browser management session with a switch using SSL.  
The web browser on your management station checks to see if the name  
to whom the certificate was issued matches the name of the web site. In  
the case of the AT-9400 Switch, the web site’s name is the switch’s IP  
address or domain name or, in the case of an enhanced stack, the master  
switch’s IP address. If the names do not match, the web browser displays  
a security warning. Of course, even if you see the security warning, you  
can close the warning prompt and still configure the switch using your web  
browser.  
Note  
If the certificate will be issued by a private or public CA, you should  
check with the CA to see if they have any rules or guidelines on  
distinguished names for the certificates they issue.  
402  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
SSL and Enhanced Stacking  
Secure Sockets Layer (SSL) is supported in an enhanced stack, but only  
when all switches in the stack are using the feature.  
When a switch’s web server is operating in HTTP, management packets  
are transmitted in plaintext. When it operates in HTTPS, management  
packets are encrypted. The web server on the AT-9400 Switch operate in  
either mode. Enhanced stacking switches that do not support SSL, such  
as the AT-8000 Series switches, use HTTP exclusively.  
A web browser management session of the switches in an enhanced stack  
cannot alternate between the different security modes during a session.  
The management session assumes that the web server mode that the  
master switch is using is the same for all the switches in the stack. As an  
example, if the master switch is using HTTPS, a web browser  
management session assumes that all the other switches in the stack are  
also using HTTPS, and it does not allow you to manage any switches  
running HTTP.  
For those networks that consist of enhanced stacking switches where  
some switches support SSL and others do not, there are two approaches  
you can take. One is to create different enhanced stacks for the different  
switches, with one enhanced stack for those switches that support SSL  
and another for those that do not. You create different enhanced stacks by  
connecting the switches with different common VLANs.  
Another workaround is to create one enhanced stack of all the switches  
and designate two master switches, where one master switch uses HTTP  
and the other HTTPS. When you need to manage those switches in the  
stack supporting SSL, you would start the management session on the  
master switch whose server mode is set to HTTPS. And when you want to  
manage those switch not supporting SSL, you would start the  
management session on the master switch whose web server is set to  
HTTP.  
Each switch in a stack must have its own key pair and certificate. They  
cannot share keys and certificates. When you start a web browser  
management session on the master switch of an enhanced stack, the  
management session uses that switch’s certificate and key pair. When you  
change to another switch in the stack, the management session starts to  
use the certificate and key pair on that switch, and so forth.  
Section IX: Management Security  
403  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 34: PKI Certificates and SSL  
Guidelines  
The guidelines for creating certificates are:  
ˆ A certificate can have only one key.  
ˆ A switch can use only those certificates that contain a key that was  
generated on the switch.  
ˆ You can create multiple certificates on a switch, but the device uses  
the certificate whose key pair has been designated as the active key  
pair for the switch’s web server.  
ˆ Most web browsers support both unsecured (plaintext) and secured  
(encrypted) operation. These modes are referred to as HTTP and  
HTTPS, respectively. If you choose to use encryption when you  
manage a switch, the web browser you use must support HTTPS.  
404  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Technical Overview  
This section describes the Secure Sockets Layer (SSL) feature, a security  
protocol that provides a secure and private TCP connection between a  
client and server.  
SSL can be used with many higher layer protocols including HTTP, File  
Transfer Protocol (FTP) and Net News Transfer Protocol (NNTP). Most  
web browsers and servers support SSL, and its most common deployment  
is for secure connections between a client and server over the Internet.  
The switch supports SSL versions 2.0 (client hello only) and 3.0 which  
were developed by Netscape, and the Internet Engineering Task Force  
(IETF) standard for SSL, known as SSL version 3.1 or Transport Layer  
Security (TLS).  
Within the Ethernet protocol stack, SSL is a Layer 4 protocol that is in  
between the HTTP and TCP protocol layers. HTTP communicates with  
SSL in the same way as with TCP. In other words, TCP processes SSL  
requests like any other protocol requesting its services.  
SSL provides a secure connection over which web pages can be  
accessed from an HTTP server. The operation of SSL is transparent to the  
end user who is accessing a web site with the following exceptions:  
ˆ The site’s URL changes from HTTP to HTTPS.  
ˆ The browser indicates that it is a secured connection by displaying an  
icon, such as a padlock icon.  
By default, HTTP and HTTPS use the separate well-known ports 80 and  
443, respectively. Secure connections over the Internet are important  
when transmitting confidential data such as credit card details or  
passwords. SSL allows the client to verify the server’s identity before  
either side sends any sensitive information. SSL also prevents a third party  
from interfering with the message because only trusted devices have  
access to the unprotected data.  
SSL Encryption SSL uses encryption to ensure the security of data transmission.  
Encryption is a process that uses an algorithm to encode data so it can  
only be accessed by a trusted device. An encrypted message remains  
confidential.  
All application data messages are authenticated by SSL with a message  
authentication code (MAC). The MAC is a checksum that is created by the  
sender and is sent as part of the encrypted message. The recipient re-  
calculates the MAC, and if the values match, the sender’s identity is  
verified. The MAC also ensures that the message has not been tampered  
with by a third party because any change to the message changes the  
MAC.  
Section IX: Management Security  
405  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 34: PKI Certificates and SSL  
SSL uses asymmetrical (Public Key) encryption to establish a connection  
between client and server, and symmetrical (Secret Key) encryption for  
the data transfer phase.  
User Verification An SSL connection has two phases: handshake and data transfer. The  
handshake initiates the SSL session, during which data is securely  
transmitted between a client and server. During the handshake, the  
following occurs:  
ˆ The client and server establish the SSL version they are to use.  
ˆ The client and server negotiate the cipher suite for the session, which  
includes encryption, authentication, and key exchange algorithms.  
ˆ The symmetrical key is exchanged.  
ˆ The client authenticates the server (optionally, the server  
authenticates the client).  
SSL messages are encapsulated by the Record Layer before being  
passed to TCP for transmission. Four types of SSL messages exist, they  
are:  
ˆ Handshake  
ˆ Change Cipher Spec  
ˆ Alert  
ˆ Application data (HTTP, FTP or NNTP)  
As discussed previously, the Handshake message initiates the SSL  
session.  
The Change Cipher Spec message informs the receiving party that all  
subsequent messages are encrypted using previously negotiated security  
options. The parties use the strongest cryptographic systems that they  
both support.  
The Alert message is used if the client or server detects an error. Alert  
messages also inform the other end that the session is about to close. In  
addition, the Alert message contains a severity rating and a description of  
the alert. For example, an alert message is sent if either party receives an  
invalid certificate or an unexpected message.  
The Application data message encapsulates the encrypted application  
data.  
Authentication Authentication is the process of ensuring that both the web site and the  
end user are genuine. In other words, they are not imposters. Both the  
server and an individual users need to be authenticated. This is especially  
important when transmitting secure data over the Internet.  
406  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
To verify the authenticity of a server, the server has a public and private  
key. The public key is given to the user.  
SSL uses certificates for authentication. A certificate binds a public key to  
a server name. A certification authority (CA) issues certificates after  
checking that a public key belongs to its claimed owner. There are several  
agencies that are trusted to issue certificates. Individual browsers have  
approved Root CAs that are built in to the browser.  
Public Key The public key infrastructure (PKI) feature is part of the switch’s suite of  
security modules, and consists of a set of tools for managing and using  
certificates. The tools that make up the PKI allow the switch to securely  
exchange public keys, while being sure of the identity of the key holder.  
Infrastructure  
The switch acts as an End Entity (EE) in a certificate-based PKI. More  
specifically, the switch can communicate with Certification Authorities  
(CAs) and Certificate Repositories to request, retrieve and verify  
certificates.The switch allows protocols running on the switch, such as  
ISAKMP, access to these certificates. The following sections of this  
chapter summarize these concepts and describe the switch’s  
implementation of them.  
Public Keys Public key encryption involves the generation of two keys for each user,  
one private and one public. Material encrypted with a private key can only  
be decrypted with the corresponding public key, and vice versa. An  
individual’s private key must be kept secret, but the public key may be  
distributed as widely as desired, because it is impossible to calculate the  
private key from the public key. The advantage of public key encryption is  
that the private key need never be exchanged, and so can be kept secure  
more easily than a shared secret key.  
Message One of the two main services provided by public key encryption is the  
exchange of encrypted messages. For example, user 1 can send a secure  
Encryption  
message to user 2 by encrypting it with user 2’s public key. Only user 2  
can decrypt it, because only user 2 has access to the corresponding  
private key.  
Digital Signatures The second main service provided by public key encryption is digital  
signing. Digital signatures both confirm the identity of the message’s  
supposed sender and protect the message from tampering. Therefore they  
provide message authentication and non-repudiation. It is very difficult for  
the signer of a message to claim that the message was corrupted, or to  
deny that it was sent.  
Both the exchange of encrypted messages and digital signatures are  
secure only if the public key used for encryption or decryption belongs to  
the message’s expected recipient. If a public key is insecurely distributed,  
it is possible a malicious agent could intercept it and replace it with the  
malicious agent’s public key (the Man-in-the-Middle attack). To prevent  
Section IX: Management Security  
407  
Download from Www.Somanuals.com. All Manuals Search And Download.  
               
Chapter 34: PKI Certificates and SSL  
this, and other attacks, PKI provides a means for secure transfer of public  
keys by linking an identity and that identity’s public key in a secure  
certificate.  
Caution  
Although a certificate binds a public key to a subject to ensure the  
public key’s security, it does not guarantee that the security of the  
associated private key has not been breached. A secure system is  
dependent upon private keys being kept secret, by protecting them  
from malicious physical and virtual access.  
Certificates A certificate is an electronic identity document. To create a certificate for a  
subject, a trusted third party (known as the Certification Authority) verifies  
the subject’s identity, binds a public key to that identity, and digitally signs  
the certificate. A person receiving a copy of the certificate can verify the  
Certification Authority’s digital signature and be sure that the public key is  
owned by the identity in it.  
The switch can generate a self-signed certificate but this should only be  
used with an SSL enabled HTTP server, or where third party trust is not  
required.  
X.509 Certificates  
The X.509 specification specifies a format for certificates. Almost all  
certificates use the X.509 version 3 format, described in RFC 2459,  
Internet X.509 Public Key Infrastructure Certificate and CRL Profile. This  
is the format which is supported by the switch.  
An X.509 v3 certificate consists of:  
ˆ A serial number, which distinguishes the certificate from all others  
issued by that issuer. This serial number is used to identify the  
certificate in a Certificate Revocation List, if necessary.  
ˆ The owner’s identity details, such as name, company and address.  
ˆ The owner’s public key, and information about the algorithm with  
which it was produced.  
ˆ The identity details of the organization which issued the certificate.  
ˆ The issuer’s digital signature and the algorithm used to produce it.  
ˆ The period for which the certificate is valid.  
ˆ Optional information is included, such as the type of application with  
which the certificate is intended to be used.  
The issuing organization’s digital signature is included in order to  
authenticate the certificate. As a result, if a certificate is tampered with  
during transmission, the tampering is detected.  
408  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
         
AT-S63 Management Software Features Guide  
Elements of a A public key infrastructure is a set of applications which manage the  
creation, retrieval, validation and storage of certificates. A PKI consists of  
the following key elements:  
Public Key  
Infrastructure  
ˆ At least one certification authority (CA), which issues and revokes  
certificates.  
ˆ At least one publicly accessible repository, which stores certificates  
and Certificate Revocation Lists.  
ˆ At least one end entity (EE), which retrieves certificates from the  
repository, validates them and uses them.  
End Entities (EE)  
End entities own public keys and may use them for encryption and digital  
signing. An entity which uses its private key to digitally sign certificates is  
not considered to be an end entity, but is a certification authority.  
The switch acts as an end entity.  
Certification Authorities  
A certification authority is an entity which issues, updates, revokes and  
otherwise manages public keys and their certificates. A CA receives  
requests for certification, validates the requester’s identity according to the  
CA’s requirements, and issues the certificate, signed with one of the CA’s  
keys. CAs may also perform the functions of end entities, in that they may  
make use of other CAs’ certificates for message encryption and  
verification of digital signatures.  
An organization may own a certification authority and issue certificates for  
use within its own networks. In addition, an organization’s certificates may  
be accepted by another network, after an exchange of certificates has  
validated a certificate for use by both parties. As an alternative, an outside  
CA may be used. The switch can interact with the CA, whether a CA is part  
of the organization or not, by sending the CA requests for certification.  
The usefulness of certificates depends on how much you trust the source  
of the certificate. You must be able to trust the issuing CA to verify  
identities reliably. The level of verification required in a given situation  
depends on the organization’s security needs.  
Section IX: Management Security  
409  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 34: PKI Certificates and SSL  
Certificate To validate a certificate, the end entity verifies the signature in the  
certificate, using the public key of the CA who issued the certificate.  
Validation  
CA Hierarchies and Certificate Chains  
It may not be practical for every individual certificate in an organization to  
be signed by one certification authority. A certification hierarchy may be  
formed, in which one CA (for example, national headquarters) is declared  
to be the root CA. This CA issues certificates to the next level down in the  
hierarchy (for example, regional headquarters), who become subordinate  
CAs and issue certificates to the next level down, and so on. A hierarchy  
may have as many levels as needed.  
Certificate hierarchies allow validation of certificates through certificate  
chains and cross-certification. If a switch X, which holds a certificate  
signed by CA X, wishes to communicate securely with a switch Y, which  
holds a certificate signed by CA Y, there are two ways in which the  
switches can validate each other’s certificates. Cross-certification occurs  
when switch X validates switch Y's CA (CA Y) by obtaining a certificate for  
switch Y's CA which has been issued by its own CA (CA X). A certificate  
chain is formed if both CA X and CA Y hold a certificate signed by a root  
CA Z, which the switches have verified out of band. Switch X can validate  
switch Y’s certificate (and vice versa) by following the chain up to CA Z.  
Root CA Certificates  
A root CA must sign its own certificate. The root CA is the most critical link  
in the certification chain, because the validity of all certificates issued by  
any CA in the hierarchy depends on the root CA’s validity. Therefore,  
every device which uses the root CA’s certificate must verify it out-of-  
band.  
Out-of-band verification involves both the owner of a certificate and the  
user who wishes to verify that certificate generating a one-way hash (a  
fingerprint) of the certificate. These two hashes must then be compared  
using at least one non-network-based communication method. Examples  
of suitable communication methods are mail, telephone, fax, or transfer by  
hand from a storage device such as a smart card or floppy disk. If the two  
hashes are the same, the certificate can be considered valid.  
Certificate A certificate may become invalid because some of the details in it change  
(for example, the address changes), because the relationship between the  
Revocation Lists  
(CRLs)  
Certification Authority (CA) and the subject changes (for example, an  
employee leaves a company), or because the associated private key is  
compromised. Every CA is required to keep a publicly accessible list of its  
certificates which have been revoked.  
410  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
         
AT-S63 Management Software Features Guide  
PKI The following sections discuss the implementation of PKI on the AT-9400  
Switch. The following topics are covered:  
Implementation  
ˆ PKI Standards  
ˆ Certificate Retrieval and Storage  
ˆ Certificate Validation  
ˆ Root CA Certificates  
PKI Standards  
The following standards are supported by the switch:  
ˆ draft-ietf-pkix-roadmap-05 — PKIX Roadmap  
ˆ RFC 1779 — A String Representation of Distinguished Names  
ˆ RFC 2459 — PKIX Certificate and CRL Profile  
ˆ RFC 2511 — PKIX Certificate Request Message Format  
ˆ PKCS #10 v1.7 — Certification Request Syntax Standard  
Certificate Retrieval and Storage  
Certificates are stored by CAs in publicly accessible repositories for  
retrieval by end entities. The following repositories used in PKI are  
commonly accessed via the following protocols: Hypertext Transfer  
Protocol (HTTP), File Transfer Protocol (FTP).  
Before the switch can use a certificate, it must be retrieved and manually  
added to the switch’s certificate database, which is stored in RAM  
memory. The switch attempts to validate the certificate, and if validation is  
successful the certificate’s public key is available for use.  
Root CA Certificate Validation  
Root CA certificates are verified out of band by comparing the certificate’s  
fingerprint (the encrypted one-way hash with which the issuing CA signs  
the certificate) with the fingerprint which the CA has supplied by a non-  
network-based method.  
Section IX: Management Security  
411  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Chapter 34: PKI Certificates and SSL  
412  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 35  
Secure Shell (SSH)  
The sections in this chapter are:  
Section IX: Management Security  
413  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 35: Secure Shell (SSH)  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
414  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
Secure management is increasingly important in modern networks, as the  
ability to easily and effectively manage switches and the requirement for  
security are two universal requirements. Switches are often remotely  
managed using remote sessions via the Telnet protocol. This method,  
however, has a serious security problem—it is only protected by plaintext  
usernames and passwords which are vulnerable to wiretapping and  
password guessing.  
The Secure Shell (SSH) protocol provides encrypted and strongly  
authenticated remote login sessions, similar to the Telnet and rlogin  
protocols, between a host running a Secure Shell server and a machine  
with a Secure Shell client.  
The AT-S63 Management Software features Secure Shell server software  
so that network managers can securely manage the switch over an  
insecure network. It offers the benefit of cryptographic authentication and  
encryption. Secure Shell can replace Telnet for remote management  
sessions.  
Section IX: Management Security  
415  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 35: Secure Shell (SSH)  
Support for SSH  
The AT-S63 implementation of the SSH protocol is compliant with the  
SSH protocol versions 1.3, 1.5, and 2.0.  
In addition, the following SSH options and features are supported:  
ˆ Inbound SSH connections (server mode) is supported.  
ˆ The following security algorithms are supported:  
128-bit Advanced Encryption Standard (AES),  
192-bit AES, and 256-bit AES  
Arcfour (RC4) security algorithm is supported.  
Triple-DES (3DES) encryption for SSH sessions is  
supported.  
ˆ RSA public keys with lengths of 512 to 2048 bits are supported. Keys  
are stored in a format compatible with other Secure Shell  
implementations, and mechanisms are provided to copy keys to and  
from the switch.  
ˆ Compression of SSH traffic.  
The following SSH options and features are not supported:  
ˆ IDEA or Blowfish encryption  
ˆ Nonencrypted Secure Shell sessions  
ˆ Tunnelling of TCP/IP traffic  
Note  
Non-encrypted Secure Shell sessions serve no purpose.  
416  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
AT-S63 Management Software Features Guide  
SSH Server  
When the SSH server is enabled, connections from SSH clients are  
accepted. When the SSH server is disabled, connections from SSH clients  
are rejected by the switch. Within the switch, the AT-S63 Management  
Software uses well-known port 22 as the SSH default port.  
Note  
If your switch is in a network that is protected by a firewall, you may  
need to configure the firewall to permit SSH connections.  
The SSH server accepts connections from configured users only.  
Acceptable users are those with a Manager or Operator login as well as  
users configured with the RADIUS and TACACS+ protocols. You can add,  
delete, and modify users with the RADIUS and TACACS+ feature.  
SSH encryption key management is implemented by the Encryption  
(ENCO) protocol. For information on how to create encryption keys, see  
Section IX: Management Security  
417  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 35: Secure Shell (SSH)  
SSH Clients  
The SSH protocol provides a secure connection between the switch and  
SSH clients. After you have configured the SSH server, you need to install  
SSH client software on your management PC. The AT-S63 Management  
Software supports both SSH1 and SSH2 clients.  
You can download client software from the Internet. Two popular SSH  
clients are PuTTY and CYGWIN. To install SSH client software, follow the  
directions from the vendor.  
After you have configured the SSH client software, you can use the client  
software to log in to the SSH server on the switch to start either a manager  
or operator management session. The SSH server can support up to one  
manager session and eight operator sessions at one time.  
418  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
SSH and Enhanced Stacking  
The AT-S63 Management Software allows for encrypted SSH  
management sessions between a management station and a master  
switch of an enhanced stack, but not with slave switches, as explained in  
this section.  
When you remotely manage a slave switch, all management  
communications are conducted through the master switch using the  
enhanced stacking feature. Management packets from your workstation  
are first directed to the master switch before being forwarded to the slave  
switch. The reverse is true as well. Management packets from a slave  
switch first pass through the master switch before reaching your  
management station.  
Enhanced stacking uses a proprietary protocol different from Telnet and  
SSH protocols. Consequently, there is no encryption between a master  
switch and a slave switch. The result is that SSH encryption only occurs  
between your workstation and the master switch, not between your  
workstation and a slave switch.  
This is illustrated in Figure 46. The figure shows an SSH management  
station that is managing a slave switch of an enhanced stack. The packets  
exchanged between the slave switch and the master switch are  
transmitted in plaintext and those exchanged between the master switch  
and the SSH management station are encrypted  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/SP Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
CLASS  
1
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
8
9
11 13 15 17 19 21 23R  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
SFP  
1000 LINK  
L/A  
SFP  
23  
SFP  
24  
/
ACT  
Slave Switch  
23  
24  
POWER  
2
4
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
Plaintext Management Packets  
(Proprietary Enhanced Stacking Protocol)  
1
3
5
7
9
11  
13  
15  
17  
19  
21  
23R  
AT-9424T/GB Gigabit Ethernet Switch  
PORT ACTIVITY  
ACT 10/100 LINK  
L/A  
D/C  
1000 LINK  
FDX  
/
/
ACT  
CLASS  
1
LASER PRODUCT  
H
D
X
/
C
O
L
TERMINAL  
PORT  
STATUS  
1
3
5
6
7
8
9
11 13 15 17 19 21 23R  
GBIC  
GBIC  
L/A  
D/C  
L/A  
D/C  
FAULT  
MASTER  
RPS  
GBIC  
1000 LINK  
L/A  
/
ACT  
Master Switch  
23  
24  
POWER  
2
4
10 12 14 16 18 20 22 24R  
2
4
6
8
10  
12  
14  
16  
18  
20  
22  
24R  
23  
24  
Encrypted Management Packets  
(SSH Protocol)  
SSH  
Management  
Workstation  
Figure 46 SSH Remote Management of a Slave Switch  
Section IX: Management Security  
419  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 35: Secure Shell (SSH)  
Because enhanced stacking does not allow for SSH encrypted  
management sessions between a management station and a slave  
switch, you configure SSH only on the master switch of a stack. Activating  
SSH on a slave switch has no affect.  
420  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
SSH Configuration Guidelines  
Here are the guidelines to configuring SSH:  
ˆ SSH requires two encryption key pairs. One key pair functions as the  
host key and the other as the server key.  
ˆ The two encryption key pairs must be of different lengths of at least  
one increment (256 bits) apart. The recommended bit size for a server  
key is 768 bits. The recommended size for the host key is 1024 bits.  
ˆ You activate and configure SSH on the master switch of an enhanced  
stack, not on slave switches.  
ˆ The AT-S63 software uses well-known port 22 as the SSH default port.  
Section IX: Management Security  
421  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Chapter 35: Secure Shell (SSH)  
General Steps to Configuring SSH  
Configuring the SSH server involves the following procedures:  
1. Create two encryption key pairs on the switch. One pair will function as  
the host key and the other the server key.  
2. Configure and activate the Secure Shell server on the switch by  
specifying the two encryption keys in the server software.  
3. Install SSH client software on your management station.  
Follow the directions provided with the client software. You can  
download SSH client software from the Internet. Two popular SSH  
clients are PuTTY and CYGWIN.  
4. Disable the Telnet server.  
Although the switch allows the SSH and Telnet servers to be enabled  
simultaneously, allowing Telnet to be enabled negates the security of  
the SSH feature.  
5. Log in to from your SSH management station using the IP address of  
the local interface on the switch.  
422  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 36  
TACACS+ and RADIUS Protocols  
This chapter describes the two authentication protocols TACACS+ and  
RADIUS. Sections in the chapter include:  
Section IX: Management Security  
423  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 36: TACACS+ and RADIUS Protocols  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
424  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
TACACS+ and RADIUS are authentication protocols that can enhance the  
security of your network. In general terms, these authentication protocols  
transfer the task of authenticating network access from a network device  
to an authentication protocol server.  
The AT-S62 software comes with TACACS+ and RADIUS client software.  
You can use the client software to add two security features to the switch.  
The first feature, described in this chapter, creates new manager accounts  
for controlling who can log onto a switch to change its parameter settings.  
The second feature is 802.1x Port-based Access Control, explained in  
which controls access to the ports on the switch by the end users and end  
nodes.  
This chapter explains the manager accounts feature. The AT-S63  
Management Software has two standard manager login accounts:  
manager and operator. The manager account lets you change a switch’s  
parameter settings while the operator account lets you view the settings,  
but not change them. Each account has its own password. The manager  
account has a default password of “friend” and the operator account has a  
default password “operator.”  
For those networks managed by just one or two network managers, you  
might not need any additional accounts. However, for larger networks  
managed by several network managers, you might want to give each  
manager his or her own management login account for a switch rather  
than have them share an account.  
This is where TACACS+ and RADIUS can be useful. TACACS+ is an  
acronym for Terminal Access Controller Access Control System. RADIUS  
is an acronym for Remote Authentication Dial In User Services. These are  
authentication protocols. You can use protocols to transfer the task of  
validating management access from the AT-9400 Switch to an  
authentication protocol server, and so be able to create your own manager  
accounts.  
With these protocols you can create a series of username and password  
combinations that define who can manage the AT-9400 Switch.  
There are three basic functions an authentication protocol provides:  
ˆ Authentication  
ˆ Authorization  
ˆ Accounting  
Section IX: Management Security  
425  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 36: TACACS+ and RADIUS Protocols  
When a network manager logs in to a switch to manage the device, the  
switch passes the username and password entered by the manager to the  
authentication protocol server. The server checks to see if the username  
and password are valid. This is referred to as authentication.  
If the combination is valid, the authentication protocol server notifies the  
switch and the switch completes the login process, allowing the manager  
to manage the switch.  
If the username and password are invalid, the authentication protocol  
server notifies the switch and the switch cancels the login.  
Authorization defines what a manager can do after logging in to a switch.  
The AT-9400 Switch supports two management levels, Manager and  
Operator. The Manager level lets you view and configure a switch’s  
parameter settings, while the Operator level only lets you view the  
settings. You must assign an authorization level to each manager  
username and password combination on the authentication server.  
The final function of an authentication protocol is keeping track of user  
activity on network devices, referred to as accounting. The AT-S63  
Management Software does not support RADIUS or TACACS+  
accounting as part of manager accounts. However, it does support  
RADIUS accounting with the 802.1x Port-based Network Access Control  
Note  
The AT-S63 Management Software does not support the two earlier  
versions of the TACACS+ protocol, TACACS and XTACACS.  
426  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Guidelines  
Here are the main steps to using the TACACS+ or RADIUS client on the  
switch.  
1. Install a TACACS+ or RADIUS server on one or more of your network  
servers or management stations. Authentication protocol server  
software is not available from Allied Telesis.  
2. Configure the TACACS+ or RADIUS authentication server.  
Here are the guidelines to follow when configuring the server for new  
manager accounts:  
To create a new manager account, enter the username and  
password combination that the network manager will use to log  
onto the switch when managing the device. The maximum length  
for a username is 38 alphanumeric characters and spaces, and the  
maximum length for a password is 16 alphanumeric characters and  
spaces.  
You must assign each account an authorization level. This differs  
depending on the server software. TACACS+ controls this through  
the sixteen (0 to 15) different levels of the Privilege attribute. A  
privilege level of “0” gives the combination Operator status. Any  
value from 1 to 15 gives the combination Manager status.  
For RADIUS, management level is controlled by the Service Type  
attribute. This attribute has 11 different values; only two apply to  
the AT-S63 Management Software. A value of Administrative for  
this attribute gives the username and password combination  
Manager access. A value of NAS Prompt assigns the combination  
Operator status.  
Note  
This manual does not explain how to configure a TACACS+ or  
RADIUS server. For instructions, refer to the documentation  
included with the server software.  
Here are the guidelines to follow when configuring the server for  
supplicant accounts for 802.1x port-based access control:  
802.1x is only supported with a RADIUS server.  
To create an account for a supplicant connected to an  
authenticator port set to the 802.1x authentication mode, enter a  
username and password combination. The maximum length for a  
username is 38 alphanumeric characters and spaces, and the  
Section IX: Management Security  
427  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 36: TACACS+ and RADIUS Protocols  
maximum length for a password is 16 alphanumeric characters  
and spaces.  
To create an account for a supplicant connected to an  
authenticator port set to the MAC address-based authentication  
mode, enter the MAC address of the node used by the supplicant  
as both its username and password. When entering the MAC  
address, do not use spaces or colons (:).  
If you are associating VLANs with supplicant accounts, refer to  
for further information.  
3. Configure the TACACS+ or RADIUS client on the switch by entering  
the IP addresses of up to three authentication servers.  
4. Activate the TACACS+ or RADIUS client on the switch.  
The switch must have a routing interface on the local subnet where the  
TACACS+ or RADIUS server is a member. The switch uses the routing  
interface’s IP address as its source address when communicating with the  
server. For background information on routing interfaces, refer to Chapter  
Note  
Prior to version 2.0.0 of the AT-S63 Management Software,  
TACACS+ or RADIUS server had to be a member of the switch’s  
management VLAN. This restriction no longer applies. The server  
can be located on any local subnet that has a routing interface.  
By default, authentication protocol is disabled in the AT-S63 Management  
Software. Before activating it, you need the following information:  
ˆ Select either TACACS+ or RADIUS as the active authentication  
protocol. Only one authentication protocol can be active on a switch at  
a time.  
ˆ Specify the IP addresses of up to three authentication servers.  
ˆ Specify the encryption keys used by the authentication servers.  
You can specify up to three RADIUS or TACACS+ servers. Specifying  
multiple servers adds redundancy to your network. For example, removing  
an authentication server from the network for maintenance does not  
prevent network managers from logging into switches if there are one or  
two other authentication servers on the network.  
When a switch receives a username and password combination from a  
network manager, it sends the combination to the first authentication  
server in its list. If the server fails to respond, the switch sends the  
combination to the next server in the list, and so on.  
428  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Note  
If no authentication server responds or if no servers have been  
defined, the AT-S63 Management Software defaults to the standard  
manager and operator accounts.  
Note  
For more information on TACACS+, refer to the RFC 1492 standard.  
For more information on RADIUS, refer to the RFC 2865 standard.  
Section IX: Management Security  
429  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 36: TACACS+ and RADIUS Protocols  
430  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 37  
Management Access Control List  
This chapter explains how to restrict Telnet and web browser management  
access to the switch with the management access control list (ACL).  
Sections in this chapter include:  
Section IX: Management Security  
431  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 37: Management Access Control List  
Supported Platforms  
This feature is supported on the following AT-9400 Switches:  
ˆ Layer 2+ Models  
AT-9408LC/SP  
AT-9424T/GB  
AT-9424T/SP  
ˆ Basic Layer 3 Models  
AT-9424T  
AT-9424Ts  
AT-9424Ts/XP  
AT-9448T/SP  
AT-9448Ts/XP  
ˆ Stack of Basic Layer 3 Switches and the AT-StackXG Stacking Module  
Not supported  
This feature can be managed from all three management interfaces:  
ˆ Command line interface  
ˆ Menus interface  
ˆ Web browser interface  
432  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Overview  
This chapter explains how to restrict remote management access of a  
switch by creating a management access control list (management ACL).  
This feature controls which management stations can remotely manage  
the device using the Telnet application protocol or a web browser.  
The switch uses the management ACL to filter the management packets  
that it receives. The switch accepts and processes only those  
management packets that meet the criteria stated in the ACL. Those  
management packets that do not meet the criteria are discarded.  
The benefit of this feature is that you can prevent unauthorized access to  
the switch by controlling which workstations are to have remote  
management access. You can even control which method, Telnet or web  
browser, that a remote manager can use.  
For example, you can create a management ACL that allows the switch to  
accept management packets only from the management stations in one  
subnet or from just one or two specific management stations.  
An access control list (ACL) is a list of one or more statements that define  
which management packets the switch accepts. Each statement, referred  
to as an access control entry (ACE), contains criteria that the switch uses  
in making the determination.  
An ACE in a management ACL is an implicit “permit” statement. This  
means that a management packet that meets the criteria of an ACE is  
processed by the switch. Consequently, the ACEs that you enter into the  
management ACL should specify which management packets you want  
the switch to process. Packets that do not meet any of the ACEs in the  
management ACL are discarded.  
Section IX: Management Security  
433  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Chapter 37: Management Access Control List  
Parts of a Management ACE  
An ACE has the following three parts:  
ˆ IP address  
ˆ Subnet mask  
ˆ Application  
IP Address You can specify the IP address of a specific management station or a  
subnet.  
Mask The mask indicates the parts of the IP address the switch should filter on.  
A binary “1” indicates the switch should filter on the corresponding bit of  
the address, while a “0” indicates that it should not. If you are filtering on a  
specific IP address, use the mask 255.255.255.255. If you are filtering on  
a subnet, the mask would depend on the address. For example, to allow  
all management stations in the subnet 149.11.11.0 to manage the switch,  
you would enter the mask 255.255.255.0.  
Application The application parameter allows you control whether the remote  
management station can manage the switch using Telnet, a web browser,  
or both. For example, you might create an ACE that states that a particular  
remote management station can only use a web browser to manage the  
switch. You can also use this option to control whether the management  
station can ping the switch.  
434  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
         
AT-S63 Management Software Features Guide  
Guidelines  
Below are guidelines for the management ACL:  
ˆ The default setting for this feature is disabled.  
ˆ A switch can have only one management ACL.  
ˆ A management ACL can have up to 256 ACEs.  
ˆ An ACE must have an IP address and mask.  
ˆ All management ACEs are implicit “permit” statements. A management  
packet that meets the criteria of an ACE is accepted by the switch.  
Consequently, the ACEs you enter into the management ACL should  
specify which management packets you want the switch to process.  
Management packets that do not meet any of the ACEs in the  
management ACL are discarded.  
ˆ A management packet that meets an ACE is immediately processed  
by the switch and is not compared against any remaining ACEs in the  
management ACL.  
ˆ The ACEs are performed in the order of their identification number,  
starting with 1.  
ˆ The management ACL does not control local management or remote  
SSH or SNMP management of a switch.  
ˆ Activating this feature without specifying any ACEs prohibits you from  
managing the switch remotely using a Telnet application or web  
browser because the switch discards all Telnet and web browser  
management packets.  
ˆ You can apply management ACLs to both master and slave switches  
in an enhanced stack. A management ACL on a master switch filters  
management packets intended for the master switch as well as those  
intended for any slave switches that you manage through the master  
switch. A management ACL applied to a slave switch filters only those  
management packets directed to the slave switch.  
Section IX: Management Security  
435  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Chapter 37: Management Access Control List  
Examples  
Following are several examples of ACEs.  
This ACE allows the management station with the IP address  
149.11.11.11 to remotely manage the switch using either the Telnet  
application protocol or a web browser, and to ping the device:  
IP Address:  
Mask:  
149.11.11.11  
255.255.255.255  
Application Type: All  
If the management ACL had only this ACE, remote management of the  
switch would be restricted to just that management station.  
This ACE permits remote Telnet and web browser management of the  
switch from all management stations in the subnet 149.11.11.0. It also  
permits the management stations to ping the switch:  
IP Address:  
Mask:  
149.11.11.0  
255.255.255.0  
Application Type: All  
This ACE permits remote web browser management of the switch from  
the subnet 149.11.11.0. The management workstations can also ping the  
device. However, since this ACE does not include Telnet management as  
an application type, that form of management is not permitted:  
IP Address:  
Mask:  
149.11.11.0  
255.255.255.0  
Application Type: Web, Ping  
A management ACL can contain multiple ACEs. The two ACEs in the next  
example allow for remote Telnet management from the subnets  
149.11.11.0 and 149.22.22.0. Web browser management and pinging the  
device are not permitted:  
ACE #1  
IP Address:  
Subnet Mask:  
Application Type:  
149.11.11.0  
255.255.255.0  
Telnet  
ACE #2  
IP Address:  
Subnet Mask:  
Application Type:  
149.22.22.0  
255.255.255.0  
Telnet  
436  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
The two ACEs in this management ACL permit remote management from  
the management station with the IP address 149.11.11.11 and all  
management stations in the subnet 149.22.22.0:  
ACE #1  
IP Address:  
Mask:  
Application Type:  
149.11.11.11  
255.255.255.255  
All  
ACE #2  
IP Address:  
Mask:  
Application Type:  
149.22.22.0  
255.255.255.0  
All  
This example allows the switch to be pinged, but not managed, by the  
management station with the IP address 149.11.11.4:  
IP Address:  
Mask:  
Application Type:  
149.11.11.4  
255.255.255.255  
Ping  
Section IX: Management Security  
437  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Chapter 37: Management Access Control List  
438  
Section IX: Management Security  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Appendix A  
AT-S63 Management Software Default  
Settings  
This appendix lists the factory default settings for the AT-S63 Management  
Software. It contains the following sections in alphabetical order:  
439  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Appendix A: AT-S63 Management Software Default Settings  
440  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Address Resolution Protocol Cache  
The following table lists the ARP cache default setting.  
ARP Cache Setting  
Default  
150 seconds  
ARP Cache Timeout  
441  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
Boot Configuration File  
The following table lists the names of the default configuration files.  
Boot Configuration File  
Stand-alone Switch  
Default  
boot.cfg  
Stack of AT-9400 Basic Layer 3 Switches  
and the AT-StackXG Stacking Module  
stack.cfg  
442  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
BOOTP Relay Agent  
The following table lists the default setting for the BOOTP relay agent.  
BOOTP Relay Agent Setting Default  
Status  
Disabled  
4
Hop Count1  
1. Hop count is not adjustable.  
443  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
Class of Service  
The following table lists the default mappings of IEEE 802.1p priority  
levels to egress port priority queues.  
IEEE 802.1p Priority Level  
Port Priority Queue  
Q1  
0
1
2
3
4
5
6
7
Q0 (lowest)  
Q2  
Q3  
Q4  
Q5  
Q6  
Q7 (highest)  
444  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Denial of Service Defenses  
The following table lists the default settings for the Denial of Service  
prevention feature.  
Denial of Service Prevention Setting  
Default  
IP Address  
Subnet Mask  
Uplink Port  
0.0.0.0  
0.0.0.0  
Highest numbered  
existing port  
SYN Flood Defense  
Smurf Defense  
Disabled  
Disabled  
Disabled  
Disabled  
Disabled  
Disabled  
Land Defense  
Teardrop Defense  
Ping of Death Defense  
IP Options Defense  
445  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
802.1x Port-Based Network Access Control  
The following table describes the 802.1x Port-based Network Access  
Control default settings.  
802.1x Port-based Network Access  
Control Settings  
Default  
Port Access Control  
Authentication Method  
Port Role  
Disabled  
RADIUS EAP  
None  
The following table lists the default settings for RADIUS accounting.  
RADIUS Accounting Settings Default  
Status  
Port  
Disabled  
1813  
Type  
Network  
Trigger Type  
Start_Stop  
Disabled  
60  
Update Status  
Update Interval  
The following table lists the default settings for an authenticator port.  
Authenticator Port Setting  
Authentication Mode  
Default  
802.1x  
Single  
Auto  
Supplicant Mode  
Port Control  
Quiet Period  
60 seconds  
30 seconds  
Enabled  
TX Period  
Reauth Enabled  
Reauth Period  
Supplicant Timeout  
Server Timeout  
Max Requests  
3600 seconds  
30 seconds  
30 seconds  
2
446  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Authenticator Port Setting  
VLAN Assignment  
Default  
Enabled  
On  
Secure VLAN  
Control Direction  
Piggyback Mode  
Guest VLAN  
Both  
Disabled  
None  
The following table lists the default settings for a supplicant port.  
Supplicant Port Setting  
Auth Period  
Default  
30 seconds  
Held Period  
Max Start  
60 seconds  
3
Start Period  
User Name  
User Password  
30 seconds  
(none)  
(none)  
447  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Appendix A: AT-S63 Management Software Default Settings  
Enhanced Stacking  
The following table lists the enhanced stacking default setting.  
Enhanced Stacking Setting  
Switch State  
Default  
Slave  
448  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Ethernet Protection Switching Ring (EPSR) Snooping  
The following table lists the EPSR default setting.  
EPSR Setting  
Default  
EPSR State  
Disabled  
449  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
Event Logs  
The following table lists the default settings for both the permanent and  
temporary event logs.  
Event Log Setting  
Default  
Status  
Enabled  
Wrap  
Full Log Action  
450  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
GVRP  
This section provides the default settings for GVRP.  
GVRP Setting  
Default  
Status  
Disabled  
Enabled  
GIP Status  
Join Timer  
Leave Timer  
20 centiseconds  
60 centiseconds  
1000 centiseconds  
Normal  
Leave All Timer  
Port Mode  
451  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
IGMP Snooping  
The following table lists the IGMP Snooping default settings.  
IGMP Snooping Setting  
IGMP Snooping Status  
Default  
Disabled  
Multicast Host Topology  
Single Host/ Port (Edge)  
260 seconds  
64  
Host/Router Timeout Interval  
Maximum IGMP Multicast Groups  
Multicast Router Ports Mode  
Auto Detect  
452  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Internet Protocol Version 4 Packet Routing  
The following table lists the IPv4 packet routing default settings.  
Packet Routing Setting  
Default  
Equal Cost Multi-path (ECMP)  
Default Route  
Enabled  
None  
Update Timer  
30 seconds  
180 seconds  
Enabled  
Invalid Timer  
Split Horizon  
Split Horizon with Poison Reverse  
Autosummarization of Routes  
Disabled  
Disabled  
Note  
The update and invalid timers are not adjustable. The switch does  
not support the IPv4 routing holddown and flush timers.  
453  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
MAC Address-based Port Security  
The following table lists the MAC address-based port security default  
settings.  
MAC Address-based Port Security  
Setting  
Default  
Security Mode  
Intrusion Action  
Participating  
MAC Limit  
Automatic (no security)  
Discard  
No  
No Limit  
454  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
MAC Address Table  
The following table lists the default setting for the MAC address table.  
MAC Address Table Setting  
MAC Address Aging Time  
Default  
300 seconds  
455  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
Management Access Control List  
The following table lists the default setting for the management access  
control list.  
Management ACL Setting  
Default  
Status  
Disabled  
456  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Manager and Operator Account  
The following table lists the manager and operator account default  
settings.  
Manager Account Setting  
Default  
Manager Login Name  
Manager Password  
manager  
friend  
Operator Login Name  
Operator Password  
operator  
operator  
10 minutes  
CLI  
Console Disconnect Timer Interval  
Console Startup Mode  
Note  
Login names and passwords are case sensitive.  
457  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
Multicast Listener Discovery Snooping  
The following table lists the MLD Snooping default settings.  
MLD Snooping Setting  
MLD Snooping Status  
Default  
Disabled  
Multicast Host Topology  
Single Host/ Port (Edge)  
260 seconds  
64  
Host/Router Timeout Interval  
Maximum MLD Multicast Groups  
Multicast Router Ports Mode  
Auto Detect  
458  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Public Key Infrastructure  
The following table lists the PKI default settings, including the generate  
enrollment request settings.  
PKI Setting  
Default  
Switch Distinguished Name  
Maximum Number of Certificates  
Request Name  
Key Pair ID  
None  
256  
None  
0
Format  
PEM  
PKCS10  
Type  
459  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
Port Settings  
The following table lists the port configuration default settings.  
Port Configuration Setting Default  
Status  
Enabled  
10/100/1000Base-T Speed  
Duplex Mode  
Auto-Negotiation  
Auto-Negotiation  
Auto-MDI/MDIX  
Disabled  
MDI/MDI-X  
Packet Filtering  
Packet Rate Limiting  
Override Priority  
Disabled  
No override  
682 cells  
Head of Line Blocking Threshold  
Back Pressure  
Disabled  
Back Pressure Threshold  
Flow Control  
7,935 cells  
Auto  
Flow Control Threshold  
7,935 cells  
460  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
RJ-45 Serial Terminal Port  
The following table lists the RJ-45 serial terminal port default settings.  
RJ-45 Serial Terminal Port Setting  
Data Bits  
Default  
8
Stop Bits  
Parity  
1
None  
None  
Flow Control  
Baud Rate  
9600 bps  
The baud rate is the only adjustable parameter on the port.  
461  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
Router Redundancy Protocol Snooping  
The following table lists the RRP Snooping default setting.  
RRP Snooping Setting  
RRP Snooping Status  
Default  
Disabled  
462  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Server-based Authentication (RADIUS and TACACS+)  
This section describes the server-based authentication, RADIUS, and  
TACACS+ client default settings.  
Server-based The following table describes the server-based authentication default  
settings.  
Authentication  
Server-based Authentication Setting  
Default  
Server-based Authentication  
Active Authentication Method  
Disabled  
TACACS+  
RADIUS Client The following table lists the RADIUS configuration default settings.  
RADIUS Configuration Setting  
Default  
Global Encryption Key  
Global Server Timeout Period  
RADIUS Server 1 Configuration  
RADIUS Server 2 Configuration  
RADIUS Server 3 Configuration  
Auth Port  
ATI  
30 seconds  
0.0.0.0  
0.0.0.0  
0.0.0.0  
1812  
Encryption Key  
Not Defined  
TACACS+ Client The following table lists the TACACS+ client configuration default settings.  
TACACS+ Client Configuration Setting  
Default  
TAC Server 1  
TAC Server 2  
TAC Server 3  
TAC Global Secret  
TAC Timeout  
0.0.0.0  
0.0.0.0  
0.0.0.0  
None  
30 seconds  
463  
Download from Www.Somanuals.com. All Manuals Search And Download.  
           
Appendix A: AT-S63 Management Software Default Settings  
Simple Network Management Protocol  
The following table describes the SNMP default settings.  
SNMP Communities Setting  
SNMP Status  
Default  
Disabled  
Disabled  
Authentication Failure Trap Status  
Community Name  
public (Read only)  
Community Name  
private (Read|Write)  
Status (public)  
Enabled  
Enabled  
No  
Status (private)  
Open Status (public)  
Open Status (private)  
No  
464  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Simple Network Time Protocol  
The following table lists the SNTP default settings.  
SNTP Setting  
Default  
System Time  
00:00:00 on January 1,  
1980  
SNTP Status  
Disabled  
0.0.0.0  
SNTP Server  
UTC Offset  
+0  
Daylight Savings Time (DST)  
Poll Interval  
Enabled  
600 seconds  
465  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
Spanning Tree Protocols (STP, RSTP, and MSTP)  
This section provides the spanning tree, STP RSTP, and MSTP, default  
settings.  
Spanning Tree The following table describes the Spanning Tree Protocol default settings  
for the switch.  
Switch Settings  
Spanning Tree Setting  
Spanning Tree Status  
Active Protocol Version  
Default  
Default  
Disabled  
RSTP  
Spanning Tree The following table describes the STP default settings.  
Protocol  
STP Setting  
Bridge Priority  
Bridge Hello Time  
Bridge Forwarding  
Bridge Max Age  
Port Cost  
32768  
2
15  
20  
Automatic -Update  
128  
Port Priority  
Rapid Spanning The following table describes the RSTP default settings.  
Tree Protocol  
RSTP Setting  
Default  
Force Version  
Bridge Priority  
Bridge Hello Time  
Bridge Forwarding  
Bridge Max Age  
Edge Port  
RSTP  
32768  
2
15  
20  
Yes  
Point-to-Point  
Port Cost  
Auto Detect  
Automatic Update  
128  
Port Priority  
466  
Download from Www.Somanuals.com. All Manuals Search And Download.  
           
AT-S63 Management Software Features Guide  
Multiple The following table lists the MSTP default settings.  
Spanning Tree  
Protocol  
MSTP Setting  
Default  
Status  
Disabled  
Force Version  
MSTP  
2
Bridge Hello Time  
Bridge Forwarding Delay  
Bridge Max Age  
Maximum Hops  
Configuration Name  
Revision Level  
15  
20  
20  
null  
0
CIST Priority  
Increment 8 (32768)  
Increment 8 (128)  
Auto Update  
Auto Detect  
Port Priority  
Port Internal Path Cost  
Port External Path Cost  
Point-to-Point  
Auto Detect  
Edge Port  
Yes  
467  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
Secure Shell Server  
The following table lists the SSH default settings.  
SSH Setting  
Default  
Status  
Disabled  
Host Key ID  
Server Key ID  
Not Defined  
Not Defined  
0 hours  
Server Key Expiry Time  
Login Timeout  
180 seconds  
22  
SSH Port Number  
The SSH port number is not adjustable.  
468  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Secure Sockets Layer  
The following table lists the SSL default settings.  
SSL Setting  
Default  
Maximum Number of Sessions  
Session Cache Timeout  
50  
300 seconds  
469  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
System Name, Administrator, and Comments Settings  
The following table describes the IP default settings.  
IP Setting  
Default  
System Name  
Administrator  
Comments  
None  
None  
None  
470  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Telnet Server  
The following table lists the Telnet server default settings.  
Telnet Server Setting  
Telnet Server  
Default  
Enabled  
23  
Telnet Port Number  
NULL Character  
Off  
The Telnet port number is not adjustable.  
471  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
Virtual Router Redundancy Protocol  
The following table lists the VRRP default setting.  
VRRP Setting  
Default  
Status  
Disabled  
472  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
VLANs  
This section provides the VLAN default settings.  
VLAN Setting  
Default VLAN Name  
Default  
Default_VLAN (all ports)  
1 (Default_VLAN)  
User Configured  
None  
Management VLAN ID  
VLAN Mode  
Uplink Port  
Ingress Filtering  
Disabled  
473  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix A: AT-S63 Management Software Default Settings  
Web Server  
The following table lists the web server default settings.  
Web Server Configuration Setting  
Status  
Default  
Enabled  
HTTP  
80  
Operating Mode  
HTTP Port Number  
HTTPS Port Number  
443  
474  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix B  
SNMPv3 Configuration Examples  
This appendix provides two examples of SNMPv3 configuration using the  
SNMPv3 Table menus and a worksheet to use as an aid when configuring  
the SNMPv3 protocol. It includes the following sections:  
475  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Appendix B: SNMPv3 Configuration Examples  
SNMPv3 Configuration Examples  
ˆ This appendix provides SNMPv3 configuration examples for the  
following types of users:  
ˆ Manager  
ˆ Operator  
In addition an SNMPv3 Configuration Table is provided to record your  
SNMPv3 configuration.  
For more information about the SNMPv3 protocol, see Chapter 19,  
SNMPv3 This section provides a sample configuration for a Manager with a User  
Name of systemadmin24. Each table is listed with its parameters.  
Manager  
Configuration  
Configure SNMPv3 User Table Menu  
User Name: systemadmin24  
Authentication Protocol: MD5  
Privacy Protocol: DES  
Storage Type: NonVolatile  
Configure SNMPv3 View Table Menu  
View Name: internet  
View Subtree OID: internet (or 1.3.6.1)  
Subtree Mask:  
View Type: Included  
Storage Type: NonVolatile  
Configure SNMPv3 Access Table  
Group Name: Managers  
Security Model: SNMPv3  
Security Level: P-Authentication and Privacy  
Read View Name: internet  
Write View Name: internet  
Notify View Name: internet  
Storage Type: NonVolatile  
476  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Configure SNMPv3 SecurityToGroup Table  
User Name:systemadmin24  
Security Model:v3  
Group Name: Managers  
Storage Type: NonVolatile  
Configure SNMPv3 Notify Table  
Notify Name: sysadminTrap  
Notify Tag: sysadminTag  
Notify Type: Trap  
Storage Type: NonVolatile  
Configure SNMPv3 Target Address Table  
Target Address Name: host451  
Target IP Address: 198.35.11.1  
UDP Port#: 162  
Timeout: 1500  
Retries: 3  
Tag List: sysadminTag  
Target Parms Name: SNMPmanagerPC  
Storage Type: NonVolatile  
Configure SNMPv3 Target Parameters Table  
Target Parameters Name:SNMPmanagerPC  
User Name:systemadmin24  
Security Model: v3  
Security Level: P-Authentication and Privacy  
Storage Type: NonVolatile  
SNMPv3 This section provides a sample configuration for an Operator with a User  
Name of nikoeng73. Because this user will only send messages to a group  
and not an SNMP host, you do not need to configure message notification  
for this user.  
Operator  
Configuration  
Configure SNMPv3 User Table Menu  
User Name: nikoeng73  
Authentication Protocol: MD5  
Privacy Protocol: None  
Storage Type: NonVolatile  
477  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Appendix B: SNMPv3 Configuration Examples  
Configure SNMPv3 View Table Menu  
View Name: internet  
View Subtree OID: 1.3.6.1 (or internet)  
Subtree Mask:  
View Type: Included  
Storage Type: NonVolatile  
Configure SNMPv3 Access Table  
Group Name: Operators  
Security Model: SNMPv3  
Security Level: Authentication  
Read View Name: internet  
Write View Name:  
Notify View Name:  
SNMPv3 This section supplies a table that you can use a worksheet when  
configuring SNMPv3. Each SNMPv3 Table is listed with its associated  
parameters.  
Worksheet  
SNMPv3 Parameters  
SNMPv3 User Table  
User Name  
Authentication Protocol  
Authentication Password  
Privacy Protocol  
Privacy Password  
Storage Type  
SNMPv3 View Table Menu  
View Name  
View Subtree OID  
Subtree Mask  
View Type  
Storage Type  
SNMPv3 Access Table Menu  
Group Name  
478  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
SNMPv3 Parameters (Continued)  
Security Model  
Security Level  
Read View Name  
Write View Name  
Notify View Name  
Storage Type  
SNMPv3 SecurityToGroup Table  
User Name  
Security Model  
Group Name  
Storage Type  
SNMPv3 Notify Table  
Notify Name  
Notify Tag  
Notify Type  
Storage Type  
SNMPv3 Target Address Table  
Target Address Name  
Target IP Address  
UDP Port  
Timeout  
Retries  
Tag List  
Target Parms Name  
Storage Type  
SNMPv3 Target Parameters Table  
Target Parameters Name  
User (Security) Name  
479  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Appendix B: SNMPv3 Configuration Examples  
SNMPv3 Parameters (Continued)  
Security Model  
Security Level  
Storage Type  
480  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Appendix C: Features and Standards  
10/100/1000Base-T Twisted Pair Ports  
IEEE 802.1d  
IEEE 802.3  
IEEE 802.3u  
IEEE 802.3ab  
IEEE 802.3u  
IEEE 802.3x  
IEEE 802.3z  
Bridging  
10Base-T  
100Base-TX  
1000Base-T  
Auto-Negotiation  
10/100 Mbps Flow Control / Backpressure  
1000 Mbps Flow Control  
Auto-MDI/MDIX  
Head of Line Blocking  
Eight Egress Queues Per Port  
Bad cable detection  
Denial of Service Defenses  
Smurf  
SYN Flood  
Teardrop  
Land  
IP Option  
Ping of Death  
Ethernet Protection Switching Ring Snooping  
Ethernet Protection Switching Ring Snooping  
482  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Fiber Optic Ports (AT-9408LC/SP Switch)  
IEEE 802.1d  
Bridging  
IEEE 802.3z  
1000Base-SX  
Head of Line Blocking  
Eight Egress Queues Per Port  
File System  
8 megabyte storage capacity  
DHCP and BOOTP Clients  
RFC 2131  
DHCP client  
RFC 951, 1542  
BOOTP client  
Internet Protocol Multicasting  
RFC 1112  
RFC 2236  
RFC 3376  
RFC 2710  
RFC 3810  
RFC 3768  
IGMP Snooping (Ver. 1.0)  
IGMP Snooping (Ver. 2.0)  
IGMP Snooping (Ver. 3.0)  
MLD Snooping (Ver. 1.0)  
MLD Snooping (Ver. 2.0)  
RRP Snooping  
Internet Protocol Version 4 Routing  
Routing Interfaces  
Static Routes  
RIP version 1  
RIP version 2  
RFC 1058  
RFC 1723  
483  
Download from Www.Somanuals.com. All Manuals Search And Download.  
         
Appendix C: Features and Standards  
RFC 826  
Address Resolution Protocol  
Equal Cost Multi-path  
Split Horizon and Split Horizon with Poison Reverse  
Autosummarization of Routes  
BOOTP Relay  
RFC 1542  
MAC Address Table  
Storage capacity of 16K entries  
Management Access and Security  
RFC 1157  
RFC 1901  
RFC 3411  
RFC 1492  
RFC 2865  
RFC 2068  
RFC 2616  
RFC 1866  
RFC 854  
SNMPv1  
SNMPv2  
SNMPv3  
TACACS+ Client  
RADIUS Client  
HTTP  
HTTPS  
HTML  
Telnet Server  
Secure Sockets Layer (SSL)  
Public Key Infrastructure (PKI)  
Encryption Keys  
Secure Shell (SSH) (Vers. 1.3, 1.5, 2.0)  
Management Access Control List  
TFTP client  
RFC 4325 (X.509)  
RFC 1350  
RFC 2030  
SNTP client  
484  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Management Access Methods  
Enhanced Stacking  
Out-of-band management (serial port)  
In-band management (over the network) using Telnet, SSH, web browser,  
and SNMP  
Management Interfaces  
Menus  
Command Line  
Web Browser  
SNMP v1, v2, & v3  
Management MIBs  
RFC 1213  
RFC 1215  
RFC 1493  
RFC 2863  
RFC 2933  
RFC 1643  
RFC 2674  
RFC 1757  
MIB-II  
TRAP MIB  
Bridge MIB  
Interface Group MIB  
IGMP  
Ethernet-like MIB  
IEEE 802.1Q MIB  
RMON 4 groups  
Allied Telesis Private MIBs  
485  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Appendix C: Features and Standards  
Port Security  
IEEE 802.1x  
Port-based Network Access Control: Supports  
multiple supplicants per port and the following  
authentication methods:  
EAP-MD5  
EAP-TLS  
EAP-TTLS  
PEAP  
RFC 2865  
RFC 2866  
RADIUS Client  
RADIUS Accounting  
MAC Address-based security  
Port Trunking and Mirroring  
IEEE 802.3ad  
Link Aggregation Control Protocol (LACP)  
Static Port Trunking  
Port Mirroring  
Spanning Tree Protocols  
IEEE 802.1D  
IEEE 802.1w  
IEEE 802.1s  
Spanning Tree Protocol  
Rapid Spanning Tree Protocol  
Multiple Spanning Tree Protocol  
System Monitoring  
RFC 3195  
Syslog Client  
Temporary Event Log (4,000 events maximum)  
Permanent Event Log (2,000 events maximum)  
Port and System Statistics  
486  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
AT-S63 Management Software Features Guide  
RFC 1757  
RFC 2386  
RMON Groups 1, 2, 3, and 9  
Traffic Control  
Quality of Service featuring:  
Layer 2, 3, and 4 criteria  
Flow Groups, Traffic Classes, and Policies  
DSCP Replacement  
802.1q Priority Replacement  
Type of Service Replacement  
Type of Service to 802.1q Priority  
Replacement  
802.1q Priority to Type of Service  
Replacement  
Maximum Bandwidth Control  
Burst Size Control  
Support for Ingress and Egress Ports  
IEEE 802.1p  
Class of Service with Strict and Weighted Round  
Robin Scheduling  
Port Access Control Lists  
Ingress and Egress Control of Broadcast, Multicast,  
and Unknown Unicast Traffic  
Ingress Packet Rate Limiting  
Virtual LANs  
IEEE 802.1Q  
Tagged VLANs  
Port-based VLANs  
Compliant and Non-compliant 802.1Q Multiple  
VLAN Modes  
Protected Ports VLANs  
487  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix C: Features and Standards  
MAC Address-based VLANs (Not supported on the  
AT-9408LC/SP, AT-9424T/GB, and AT-9424T/SP  
switches.)  
IEEE 802.3ac  
IEEE 802.1P  
VLAN Tag Frame Extension  
GARP VLAN Registration Protocol  
Virtual Router Redundancy Protocol  
RFC 3768  
Virtual Router Redundancy Protocol  
488  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Appendix D  
MIB Objects  
This appendix lists the SNMP MIB objects in the private Allied Telesis  
MIBs that apply to the AT-S63 Management Software and the AT-9400  
Switch. Sections in the appendix include:  
The Allied Telesis MIB files for the AT-9400 Switch are:  
ˆ atiStackSwitch.mib (version 2.31)  
ˆ atiStackInfo.mib (version 1.3)  
The MIB files are available from the Allied Telesis web site. Objects in the  
private MIBs have the prefix “1.3.6.1.4.1.207.”  
Note  
The AT-9400 Switch supports only a limited number of management  
functions from SNMP.  
489  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Appendix D: MIB Objects  
Access Control Lists  
Table 31. Access Control Lists (AtiStackSwitch MIB)  
Object Name  
atiStkSwACLConfigTable  
OID  
1.3.6.1.4.1.207.8.17.9.1  
atiStkSwACLConfigEntry  
atiStkSwACLModuleId  
atiStkSwACLId  
1.3.6.1.4.1.207.8.17.9.1.1  
1.3.6.1.4.1.207.8.17.9.1.1.1  
1.3.6.1.4.1.207.8.17.9.1.1.2  
1.3.6.1.4.1.207.8.17.9.1.1.3  
1.3.6.1.4.1.207.8.17.9.1.1.4  
1.3.6.1.4.1.207.8.17.9.1.1.5  
1.3.6.1.4.1.207.8.17.9.1.1.6  
1.3.6.1.4.1.207.8.17.9.1.1.7  
atiStkSwACLDescription  
atiStkSwACLAction  
atiStkSwACLClassifierList  
atiStkSwACLPortList  
atiStkSwACLRowStatus  
490  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Class of Service  
Table 32. CoS Scheduling (AtiStackSwitch MIB)  
Object Name  
OID  
atiSwQoSGroup  
1.3.6.1.4.1.207.8.17.7  
1.3.6.1.4.1.207.8.17.7.1  
1.3.6.1.4.1.207.8.17.7.2  
atiStkSwQoSGroupNumberOfQueues  
atiStkSwQoSGroupSchedulingMode  
Table 33. CoS Priority to Egress Queue Mappings (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwQoSGroupCoSToQueueTable  
atiStkSwQoSGroupCoSToQueueEntry  
atiStkSwQoSGroupCoSPriority  
1.3.6.1.4.1.207.8.17.7.3  
1.3.6.1.4.1.207.8.17.7.3.1  
1.3.6.1.4.1.207.8.17.7.3.1.1  
1.3.6.1.4.1.207.8.17.7.3.1.2  
atiStkSwQoSGroupCoSQueue  
Table 34. CoS Packet Weights of Egress Queues (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwQoSGroupQueueToWeightTable  
AtiStkSwQoSGroupQueueToWeightEntry  
atiStkSwQoSGroupQueue  
1.3.6.1.4.1.207.8.17.7.4  
1.3.6.1.4.1.207.8.17.7.4.1  
1.3.6.1.4.1.207.8.17.7.4.1.1  
1.3.6.1.4.1.207.8.17.7.4.1.2  
atiStkSwQoSGroupQueueWeight  
Table 35. CoS Port Settings (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwQoSGroupPortCoSPriorityTable  
atiStkSwQoSGroupPortCoSPriorityEntry  
atiStkSwQoSGroupPortCoSPriorityModuleId  
atiStkSwQoSGroupPortCoSPriorityPortId  
atiStkSwQoSGroupPortCoSPriorityPriority  
1.3.6.1.4.1.207.8.17.7.8  
1.3.6.1.4.1.207.8.17.7.8.1  
1.3.6.1.4.1.207.8.17.7.8.1.1  
1.3.6.1.4.1.207.8.17.7.8.1.2  
1.3.6.1.4.1.207.8.17.7.8.1.3  
atiStkSwQoSGroupPortCoSPriorityOverridePriority 1.3.6.1.4.1.207.8.17.7.8.1.4  
491  
Download from Www.Somanuals.com. All Manuals Search And Download.  
         
Appendix D: MIB Objects  
Date, Time, and SNTP Client  
Table 36. Date, Time, and SNTP Client (AtiStackSwitch MIB)  
Object Name  
atiStkSysSystemTimeConfig  
OID  
1.3.6.1.4.1.207.8.17.1.5  
1.3.6.1.4.1.207.8.17.1.5.1  
1.3.6.1.4.1.207.8.17.1.5.2  
1.3.6.1.4.1.207.8.17.1.5.3  
1.3.6.1.4.1.207.8.17.1.5.4  
1.3.6.1.4.1.207.8.17.1.5.5  
1.3.6.1.4.1.207.8.17.1.5.6  
1.3.6.1.4.1.207.8.17.1.5.7  
1.3.6.1.4.1.207.8.17.1.5.8  
atiStkSwSysCurrentTime  
atiStkSwSysCurrentDate  
atiStkSwSysSNTPStatus  
atiStkSwSysSNTPServerIPAddress  
atiStkSwSysSNTPUTCOffset  
atiStkSwSysSNTPDSTStatus  
atiStkSwSysSNTPPollingInterval  
atiStkSwSysSNTPLastDelta  
492  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Denial of Service Defenses  
Table 37. LAN Address and Subnet Mask (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkDOSConfig  
1.3.6.1.4.1.207.8.17.2.6  
1.3.6.1.4.1.207.8.17.2.6.1  
1.3.6.1.4.1.207.8.17.2.6.2  
atiStkDOSConfigLANIpAddress  
atiStkDOSConfigLANSubnetMask  
Table 38. Denial of Service Defenses (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkPortDOSAttackConfigTable  
atiStkPortDOSAttackConfigEntry  
atiStkPortDOSAttackType  
1.3.6.1.4.1.207.8.17.2.6.3  
1.3.6.1.4.1.207.8.17.2.6.3.1  
1.3.6.1.4.1.207.8.17.2.6.3.1.1  
1.3.6.1.4.1.207.8.17.2.6.3.1.2  
1.3.6.1.4.1.207.8.17.2.6.3.1.3  
1.3.6.1.4.1.207.8.17.2.6.3.1.4  
atiStkPortDOSAttackActionStatus  
atiStkPortDOSAttackMirrorPort  
atiStkPortDOSAttackMirrorPortStatus  
493  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Appendix D: MIB Objects  
Enhanced Stacking  
Table 39. Switch Mode and Discovery (AtiStackInfo MIB)  
Object Name  
atiswitchEnhancedStackingInfo  
OID  
1.3.6.1.4.1.207.8.16.1  
1.3.6.1.4.1.207.8.16.1.1  
1.3.6.1.4.1.207.8.16.1.2  
1.3.6.1.4.1.207.8.16.1.3  
atiswitchEnhStackMode  
atiswitchEnhStackDiscover  
atiswitchEnhStackRemoteNumber  
Table 40. Switches of an Enhanced Stack (AtiStackInfo MIB)  
Object Name  
OID  
atiswitchEnhStackTable  
1.3.6.1.4.1.207.8.16.1.4  
atiswitchEnhStackEntry  
1.3.6.1.4.1.207.8.16.1.4.1  
1.3.6.1.4.1.207.8.16.1.4.1.1  
1.3.6.1.4.1.207.8.16.1.4.1.2  
1.3.6.1.4.1.207.8.16.1.4.1.3  
1.3.6.1.4.1.207.8.16.1.4.1.4  
1.3.6.1.4.1.207.8.16.1.4.1.5  
1.3.6.1.4.1.207.8.16.1.4.1.6  
1.3.6.1.4.1.207.8.16.1.4.1.7  
atiswitchEnhStackSwId  
atiswitchEnhStackSwMacAddr  
atiswitchEnhStackSwName  
atiswitchEnhStackSwMode  
atiswitchEnhStackSwSoftwareVersion  
atiswitchEnhStackSwModel  
atiswitchEnhStackConnect  
494  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
GVRP  
Table 41. GVFP Switch Configuration (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwGVRPConfig  
1.3.6.1.4.1.207.8.17.3.6  
1.3.6.1.4.1.207.8.17.3.6.1  
1.3.6.1.4.1.207.8.17.3.6.2  
1.3.6.1.4.1.207.8.17.3.6.3  
1.3.6.1.4.1.207.8.17.3.6.4  
1.3.6.1.4.1.207.8.17.3.6.5  
atiStkSwGVRPStatus  
atiStkSwGVRPGIPStatus  
atiStkSwGVRPJoinTimer  
atiStkSwGVRPLeaveTimer  
atiStkSwGVRPLeaveAllTimer  
Table 42. GVRP Port Configuration (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwGVRPPortConfigTable  
atiStkSwGVRPPortConfigEntry  
atiStkSwGVRPPortConfigModuleId  
atiStkSwGVRPPortConfigPortId  
atiStkSwGVRPPortConfigStatus  
1.3.6.1.4.1.207.8.17.3.7  
1.3.6.1.4.1.207.8.17.3.7.1  
1.3.6.1.4.1.207.8.17.3.7.1.1  
1.3.6.1.4.1.207.8.17.3.7.1.2  
1.3.6.1.4.1.207.8.17.3.7.1.3  
Table 43. GVRP Counters (AtiStackSwitch MIB)  
Object Name  
atiStkSwGVRPCountersTable  
OID  
1.3.6.1.4.1.207.8.17.3.8  
atiStkSwGVRPCountersEntry  
1.3.6.1.4.1.207.8.17.3.8.1  
1.3.6.1.4.1.207.8.17.3.8.1.1  
1.3.6.1.4.1.207.8.17.3.8.1.2  
1.3.6.1.4.1.207.8.17.3.8.1.3  
1.3.6.1.4.1.207.8.17.3.8.1.4  
1.3.6.1.4.1.207.8.17.3.8.1.5  
1.3.6.1.4.1.207.8.17.3.8.1.6  
1.3.6.1.4.1.207.8.17.3.8.1.7  
atiStkSwGVRPCountersModuleId  
atiStkSwGVRPCountersGARPRxPkt  
atiStkSwGVRPCountersInvalidGARPRxPkt  
atiStkSwGVRPCountersGARPTxPkt  
atiStkSwGVRPCountersGARPTxDisabled  
atiStkSwGVRPCountersPortNotSending  
atiStkSwGVRPCountersGARPDisabled  
495  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Appendix D: MIB Objects  
Table 43. GVRP Counters (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwGVRPCountersPortNotListening  
1.3.6.1.4.1.207.8.17.3.8.1.8  
1.3.6.1.4.1.207.8.17.3.8.1.9  
1.3.6.1.4.1.207.8.17.3.8.1.10  
1.3.6.1.4.1.207.8.17.3.8.1.11  
1.3.6.1.4.1.207.8.17.3.8.1.12  
1.3.6.1.4.1.207.8.17.3.8.1.13  
1.3.6.1.4.1.207.8.17.3.8.1.14  
1.3.6.1.4.1.207.8.17.3.8.1.15  
1.3.6.1.4.1.207.8.17.3.8.1.16  
1.3.6.1.4.1.207.8.17.3.8.1.17  
1.3.6.1.4.1.207.8.17.3.8.1.18  
1.3.6.1.4.1.207.8.17.3.8.1.19  
1.3.6.1.4.1.207.8.17.3.8.1.20  
1.3.6.1.4.1.207.8.17.3.8.1.21  
1.3.6.1.4.1.207.8.17.3.8.1.22  
1.3.6.1.4.1.207.8.17.3.8.1.23  
1.3.6.1.4.1.207.8.17.3.8.1.24  
1.3.6.1.4.1.207.8.17.3.8.1.25  
1.3.6.1.4.1.207.8.17.3.8.1.26  
atiStkSwGVRPCountersInvalidPort  
atiStkSwGVRPCountersInvalidProtocol  
atiStkSwGVRPCountersInvalidFormat  
atiStkSwGVRPCountersDatabaseFull  
atiStkSwGVRPCountersRxMsgLeaveAll  
atiStkSwGVRPCountersRxMsgJoinEmpty  
atiStkSwGVRPCountersRxMsgJoinIn  
atiStkSwGVRPCountersRxMsgLeaveEmpty  
atiStkSwGVRPCountersRxMsgLeaveIn  
atiStkSwGVRPCountersRxMsgEmpty  
atiStkSwGVRPCountersRxMsgBadMsg  
atiStkSwGVRPCountersRxMsgBadAttribute  
atiStkSwGVRPCountersTxMsgLeaveAll  
atiStkSwGVRPCountersTxMsgJoinEmpty  
atiStkSwGVRPCountersTxMsgJoinIn  
atiStkSwGVRPCountersTxMsgLeaveEmpty  
atiStkSwGVRPCountersTxMsgLeaveIn  
atiStkSwGVRPCountersTxMsgEmpty  
496  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
MAC Address Table  
Table 44. MAC Address Table (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwMacAddr2VlanTable  
atiStkSwMacAddr2VlanEntry  
atiStkSwMacAddress  
1.3.6.1.4.1.207.8.17.3.3  
1.3.6.1.4.1.207.8.17.3.3.1  
1.3.6.1.4.1.207.8.17.3.3.1.1  
1.3.6.1.4.1.207.8.17.3.3.1.2  
1.3.6.1.4.1.207.8.17.3.3.1.3  
1.3.6.1.4.1.207.8.17.3.3.1.4  
1.3.6.1.4.1.207.8.17.3.3.1.5  
1.3.6.1.4.1.207.8.17.3.3.1.6  
atiStkSwMacAddrVlanId  
atiStkSwMacAddrVlanName  
atiStkSwMacAddrModuleId  
atiStkSwMacAddrPortId  
atiStkSwMacAddrPortList  
Table 45. Static MAC Address Table (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwMacAddrGroup  
1.3.6.1.4.1.207.8.17.4  
atiStkSwStaticMacAddrEntry  
atiStkSwStaticMacAddress  
1.3.6.1.4.1.207.8.17.4.1.1  
1.3.6.1.4.1.207.8.17.4.1.1.1  
1.3.6.1.4.1.207.8.17.4.1.1.2  
1.3.6.1.4.1.207.8.17.4.1.1.3  
1.3.6.1.4.1.207.8.17.4.1.1.4  
1.3.6.1.4.1.207.8.17.4.1.1.5  
1.3.6.1.4.1.207.8.17.4.1.1.6  
atiStkSwStaticMacAddrVlanId  
atiStkSwStaticMacAddrModuleId  
atiStkSwStaticMacAddrPortId  
atiStkSwStaticMacAddrPortList  
atiStkSwStaticMacAddrEntryStatus  
497  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Appendix D: MIB Objects  
Management Access Control List  
Table 46. Management Access Control List Status (AtiStackSwitch MIB)  
Object Name  
atiStkSwSysMgmtACLGroup  
atiStkSwSysMgmtACLStatus  
OID  
1.3.6.1.4.1.207.8.17.1.7  
1.3.6.1.4.1.207.8.17.1.7.1  
Table 47. Management Access Control List Entries (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwSysMgmtACLConfigTable  
atiStkSwSysMgmtACLConfigEntry  
atiStkSwSysMgmtACLConfigModuleId  
atiStkSwSysMgmtACLConfigId  
1.3.6.1.4.1.207.8.17.1.7.2  
1.3.6.1.4.1.207.8.17.1.7.2.1  
1.3.6.1.4.1.207.8.17.1.7.2.1.1  
1.3.6.1.4.1.207.8.17.1.7.2.1.2  
1.3.6.1.4.1.207.8.17.1.7.2.1.3  
1.3.6.1.4.1.207.8.17.1.7.2.1.4  
1.3.6.1.4.1.207.8.17.1.7.2.1.5  
1.3.6.1.4.1.207.8.17.1.7.2.1.6  
atiStkSwSysMgmtACLConfigIpAddr  
atiStkSwSysMgmtACLConfigMask  
atiStkSwSysMgmtACLConfigApplication  
atiStkSwSysMgmtACLConfigRowStatus  
498  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
AT-S63 Management Software Features Guide  
Miscellaneous  
Table 48. System Reset (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwSysGroup  
1.3.6.1.4.1.207.8.17.1  
1.3.6.1.4.1.207.8.17.1.1  
1.3.6.1.4.1.207.8.17.1.1.1  
atiStkSwSysConfig  
atiStkSwSysReset  
Table 49. Local Interface (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwSysGroup  
atiStkSwSysConfig  
1.3.6.1.4.1.207.8.17.1  
1.3.6.1.4.1.207.8.17.1.1  
1.3.6.1.4.1.207.8.17.1.1.2  
1.3.6.1.4.1.207.8.17.1.1.3  
1.3.6.1.4.1.207.8.17.1.1.4  
1.3.6.1.4.1.207.8.17.1.1.5  
atiStkSwSysIpAddress  
atiStkSwSysSubnetMask  
atiStkSwSysGateway  
atiStkSwSysIpAddressStatus  
Table 50. Saving the Configuration and Returning to Default Settings (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwSysGroup  
1.3.6.1.4.1.207.8.17.1  
atiStkSwSysConfig  
atiStkSwSysAction  
1.3.6.1.4.1.207.8.17.1.1  
1.3.6.1.4.1.207.8.17.1.1.11  
499  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
Appendix D: MIB Objects  
Port Mirroring  
Table 51. Port Mirroring (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwPortMirroringConfig  
1.3.6.1.4.1.207.8.17.2.2  
1.3.6.1.4.1.207.8.17.2.2.1  
1.3.6.1.4.1.207.8.17.2.2.4  
1.3.6.1.4.1.207.8.17.2.2.5  
1.3.6.1.4.1.207.8.17.2.2.6  
1.3.6.1.4.1.207.8.17.2.2.7  
atiStkSwPortMirroringState  
atiStkSwPortMirroringDestinationModuleId  
atiStkSwPortMirroringDestinationPortId  
atiStkSwPortMirroringSourceRxList  
atiStkSwPortMirroringSourceTxList  
500  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Quality of Service  
Table 52. Flow Groups (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwQosFlowGrpTable  
1.3.6.1.4.1.207.8.17.7.5  
atiStkSwQosFlowGrpEntry  
1.3.6.1.4.1.207.8.17.7.5.1  
1.3.6.1.4.1.207.8.17.7.5.1.1  
1.3.6.1.4.1.207.8.17.7.5.1.2  
1.3.6.1.4.1.207.8.17.7.5.1.3  
1.3.6.1.4.1.207.8.17.7.5.1.4  
1.3.6.1.4.1.207.8.17.7.5.1.5  
1.3.6.1.4.1.207.8.17.7.5.1.6  
1.3.6.1.4.1.207.8.17.7.5.1.7  
1.3.6.1.4.1.207.8.17.7.5.1.8  
1.3.6.1.4.1.207.8.17.7.5.1.9  
1.3.6.1.4.1.207.8.17.7.5.1.10  
1.3.6.1.4.1.207.8.17.7.5.1.11  
atiStkSwQosFlowGrpModuleId  
atiStkSwQosFlowGrpId  
atiStkSwQosFlowGrpDescription  
atiStkSwQosFlowGrpDSCPValue  
atiStkSwQosFlowGrpPriority  
atiStkSwQosFlowGrpRemarkPriority  
atiStkSwQosFlowGrpTos  
atiStkSwQosFlowGrpTosToPriority  
atiStkSwQosFlowGrpPriorityToTos  
atiStkSwQosFlowGrpClassifierList  
atiStkSwQosFlowGrpRowStatus  
Table 53. Traffic Classes (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwQosTrafficClassTable  
1.3.6.1.4.1.207.8.17.7.6  
atiStkSwQosTrafficClassEntry  
1.3.6.1.4.1.207.8.17.7.6.1  
1.3.6.1.4.1.207.8.17.7.6.1.1  
1.3.6.1.4.1.207.8.17.7.6.1.2  
1.3.6.1.4.1.207.8.17.7.6.1.3  
1.3.6.1.4.1.207.8.17.7.6.1.4  
1.3.6.1.4.1.207.8.17.7.6.1.5  
1.3.6.1.4.1.207.8.17.7.6.1.6  
1.3.6.1.4.1.207.8.17.7.6.1.7  
1.3.6.1.4.1.207.8.17.7.6.1.8  
atiStkSwQosTrafficClassModuleId  
atiStkSwQosTrafficClassId  
atiStkSwQosTrafficClassDescription  
atiStkSwQosTrafficClassExceedAction  
atiStkSwQosTrafficClassExceedRemarkValue  
atiStkSwQosTrafficClassDSCPValue  
atiStkSwQosTrafficClassMaxBandwidth  
atiStkSwQosTrafficClassBurstSize  
501  
Download from Www.Somanuals.com. All Manuals Search And Download.  
     
Appendix D: MIB Objects  
Table 53. Traffic Classes (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwQosTrafficClassClassPriority  
1.3.6.1.4.1.207.8.17.7.6.1.9  
1.3.6.1.4.1.207.8.17.7.6.1.10  
1.3.6.1.4.1.207.8.17.7.6.1.11  
1.3.6.1.4.1.207.8.17.7.6.1.12  
1.3.6.1.4.1.207.8.17.7.6.1.13  
1.3.6.1.4.1.207.8.17.7.6.1.14  
1.3.6.1.4.1.207.8.17.7.6.1.15  
atiStkSwQosTrafficClassRemarkPriority  
atiStkSwQosTrafficClassToS  
atiStkSwQosTrafficClassMoveToSToPriority  
atiStkSwQosTrafficClassMovePriorityToToS  
atiStkSwQosTrafficClassFlowGroupList  
atiStkSwQosTrafficClassStatus  
Table 54. Policies (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwQosPolicyTable  
1.3.6.1.4.1.207.8.17.7.7  
atiStkSwQosPolicyEntry  
1.3.6.1.4.1.207.8.17.7.7.1  
1.3.6.1.4.1.207.8.17.7.7.1.1  
1.3.6.1.4.1.207.8.17.7.7.1.2  
1.3.6.1.4.1.207.8.17.7.7.1.3  
1.3.6.1.4.1.207.8.17.7.7.1.4  
1.3.6.1.4.1.207.8.17.7.7.1.5  
1.3.6.1.4.1.207.8.17.7.7.1.6  
1.3.6.1.4.1.207.8.17.7.7.1.7  
1.3.6.1.4.1.207.8.17.7.7.1.8  
1.3.6.1.4.1.207.8.17.7.7.1.9  
1.3.6.1.4.1.207.8.17.7.7.1.10  
1.3.6.1.4.1.207.8.17.7.7.1.11  
1.3.6.1.4.1.207.8.17.7.7.1.12  
1.3.6.1.4.1.207.8.17.7.7.1.13  
1.3.6.1.4.1.207.8.17.7.7.1.14  
atiStkSwQosPolicyModuleId  
atiStkSwQosPolicyId  
atiStkSwQosPolicyDescription  
atiStkSwQosPolicyRemarkDSCP  
atiStkSwQosPolicyDSCPValue  
atiStkSwQosPolicyDSCPValue  
atiStkSwQosPolicyMoveToSToPriority  
atiStkSwQosPolicyMovePriorityToToS  
atiStkSwQosPolicySendToMirrorPort  
atiStkSwQosPolicyClassList  
atiStkSwQosPolicyRedirectPort  
atiStkSwQosPolicyIngressPortList  
atiStkSwQosPolicyEgressPortList  
atiStkSwQosPolicyRowStatus  
502  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
AT-S63 Management Software Features Guide  
Port Configuration and Status  
Table 55. Port Configuration and Status (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwPortConfigTable  
atiStkPortConfigEntry  
atiStkSwModuleId  
1.3.6.1.4.1.207.8.17.2.1  
1.3.6.1.4.1.207.8.17.2.1.1  
1.3.6.1.4.1.207.8.17.2.1.1.1  
1.3.6.1.4.1.207.8.17.2.1.1.2  
1.3.6.1.4.1.207.8.17.2.1.1.3  
1.3.6.1.4.1.207.8.17.2.1.1.4  
1.3.6.1.4.1.207.8.17.2.1.1.5  
1.3.6.1.4.1.207.8.17.2.1.1.6  
1.3.6.1.4.1.207.8.17.2.1.1.7  
1.3.6.1.4.1.207.8.17.2.1.1.8  
1.3.6.1.4.1.207.8.17.2.1.1.9  
1.3.6.1.4.1.207.8.17.2.1.1.10  
1.3.6.1.4.1.207.8.17.2.1.1.11  
1.3.6.1.4.1.207.8.17.2.1.1.12  
1.3.6.1.4.1.207.8.17.2.1.1.13  
1.3.6.1.4.1.207.8.17.2.1.1.14  
1.3.6.1.4.1.207.8.17.2.1.1.15  
1.3.6.1.4.1.207.8.17.2.1.1.16  
atiStkSwPortId  
atiStkSwPortName  
atiStkSwPortState  
atiStkSwPortLinkState  
atiStkSwPortNegotiation  
atiStkSwPortSpeed  
atiStkSwPortDuplexStatus  
atiStkSwPortFlowControl  
atiStkSwPortBackPressure  
atiStkSwPortPriority  
atiStkSwPortBroadcastProcessing  
atiStkSwPortMDIO  
atiStkSwPortHOLLimit  
atiStkSwPortBackPressureLimit  
atiStkSwPortSTPState  
503  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix D: MIB Objects  
Spanning Tree  
Table 56. Spanning Tree (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwSysConfig  
1.3.6.1.4.1.207.8.17.1.1  
1.3.6.1.4.1.207.8.17.1.1.9  
1.3.6.1.4.1.207.8.17.1.1.10  
atiStkSwSysSpanningTreeStatus  
atiStkSwSysSpanningTreeVersion  
504  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
AT-S63 Management Software Features Guide  
Static Port Trunk  
Table 57. Static Port Trunks (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwStaticTrunkTable  
atiStkSwStaticTrunkEntry  
atiStkSwStaticTrunkModuleId  
atiStkSwStaticTrunkIndex  
atiStkSwStaticTrunkId  
1.3.6.1.4.1.207.8.17.8.1  
1.3.6.1.4.1.207.8.17.8.1.1  
1.3.6.1.4.1.207.8.17.8.1.1.1  
1.3.6.1.4.1.207.8.17.8.1.1.2  
1.3.6.1.4.1.207.8.17.8.1.1.3  
1.3.6.1.4.1.207.8.17.8.1.1.4  
1.3.6.1.4.1.207.8.17.8.1.1.5  
1.3.6.1.4.1.207.8.17.8.1.1.6  
1.3.6.1.4.1.207.8.17.8.1.1.7  
1.3.6.1.4.1.207.8.17.8.1.1.8  
atiStkSwStaticTrunkName  
atiStkSwStaticTrunkMethod  
atiStkSwStaticTrunkPortList  
atiStkSwStaticTrunkStatus  
atiStkSwStaticTrunkRowStatus  
505  
Download from Www.Somanuals.com. All Manuals Search And Download.  
   
Appendix D: MIB Objects  
VLANs  
The objects in Table 58 display the specifications of the Default_VLAN.  
Table 58. VLAN Table (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwVlanConfigTable  
atiStkSwVlanConfigEntry  
atiStkSwVlanId  
1.3.6.1.4.1.207.8.17.3.1  
1.3.6.1.4.1.207.8.17.3.1.1  
1.3.6.1.4.1.207.8.17.3.1.1.1  
1.3.6.1.4.1.207.8.17.3.1.1.2  
1.3.6.1.4.1.207.8.17.3.1.1.3  
1.3.6.1.4.1.207.8.17.3.1.1.4  
1.3.6.1.4.1.207.8.17.3.1.1.19  
1.3.6.1.4.1.207.8.17.3.1.1.20  
atiStkSwVlanName  
atiStkSwVlanTaggedPortListModule1  
atiStkSwVlanUntaggedPortListModule1  
atiStkSwVlanConfigEntryStatus  
atiStkSwVlanActualUntaggedPortListModule1  
The objects in Table 59 display the names and VIDs of all the VLANs on a  
switch, but not the VLAN ports.  
Table 59. VLAN Table (AtiStackSwitch MIB)  
Object Name  
atiStkSwPort2VlanTable  
OID  
1.3.6.1.4.1.207.8.17.3.2  
1.3.6.1.4.1.207.8.17.3.2.1  
1.3.6.1.4.1.207.8.17.3.2.1.1  
1.3.6.1.4.1.207.8.17.3.2.1.2  
atiStkSwPort2VlanEntry  
atiStkSwPortVlanId  
atiStkSwPortVlanName  
Table 60. VLAN Mode and Uplink Port (AtiStackSwitch MIB)  
Object Name  
OID  
atiStkSwVlanGroup  
1.3.6.1.4.1.207.8.17.3  
1.3.6.1.4.1.207.8.17.3.4  
1.3.6.1.4.1.207.8.17.3.5  
atiStkSwVlanMode  
atiStkSwVlanUplinkVlanPort  
506  
Download from Www.Somanuals.com. All Manuals Search And Download.  
       
AT-S63 Management Software Features Guide  
Table 61. PVID Table (AtiStackSwitch MIB)  
Object Name  
atiStkSwPort2VlanTable  
OID  
1.3.6.1.4.1.207.8.17.3.2  
1.3.6.1.4.1.207.8.17.3.2.1  
1.3.6.1.4.1.207.8.17.3.2.1.1  
1.3.6.1.4.1.207.8.17.3.2.1.2  
atiStkSwPort2VlanEntry  
atiStkSwPortVlanId  
atiStkSwPortVlanName  
507  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Appendix D: MIB Objects  
508  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Index  
Numerics  
B
802.1x Port-based Network Access Control  
backup switches in Virtual Router Redundancy Protocol  
(VRRP) 341  
BOOTP relay agent  
A
C
access control entries (ACE)  
CA. See certification authority (CA)  
certificates  
access control lists  
certification authority (CA)  
ACE. See access control entry (ACE)  
ACL. See access control lists  
Address Resolution Protocol (ARP) table  
CIST. See Common and Internal Spanning Tree  
Class of Service (CoS)  
classifiers  
AT-S63 Management Software  
AT-S63 software updates  
See also RADIUS, TACACS+  
509  
Download from Www.Somanuals.com. All Manuals Search And Download.  
 
Index  
enhanced stacking  
Common and Internal Spanning Tree (CIST)  
community names  
configuration files. See boot configuration files  
control messages, Ethernet Protection Switching Ring  
(EPSR) snooping 189  
EPSR snooping. See Ethernet Protection Switching Ring  
(EPSR) snooping  
CoS. See Class of Service (CoS)  
CRL. See certificate revocation list (CRL)  
Ethernet Protection Switching Ring (EPSR) snooping  
D
default route  
denial of service defenses  
event logs  
F
G
destination MAC addresses  
GARP VLAN Registration Protocol (GVRP)  
DoS. See denial of service defenses  
GARP. See Generic Attribute Registration Protocol (GARP)  
GVRP. See GARP VLAN Registration Protocol (GVRP)  
H
E
encryption keys  
I
IGMP snooping. See Internet Group Management Protocol  
(IGMP) snooping  
510  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
Internet Group Management Protocol (IGMP) snooping  
management access control list  
Internet Protocol version 4 routing  
see also routing interfaces, Routing Information Proto-  
col (RIP), static routes  
master switch  
See also MAC address-based port security  
MIB tree  
K
MIBs  
L
MLD snooping. See Multicast Listener Discovery (MLD)  
snooping  
LACP. See Link Aggregation Control Protocol (LACP) port  
trunk  
Link Aggregation Control Protocol (LACP) port trunk  
MSTI. See Multiple Spanning Tree Instances (MSTI)  
MSTP. See Multiple Spanning Tree Protocol (MSTP)  
Multicast Listener Discovery (MLD) snooping  
load distribution methods  
Multiple Spanning Tree Protocol (MSTP)  
M
MAC address-based port security  
N
MAC address-based VLANs  
511  
Download from Www.Somanuals.com. All Manuals Search And Download.  
Index  
O
PVID. See Port VLAN identifier (PVID)  
P
Q
PKI. See Public Key Infrastructure (PKI)  
policies  
QoS. See Quality of Service (QoS)  
Quality of Service (QoS)  
See also traffic classes; flow groups; policies  
R
RADIUS  
port mirror  
port monitoring in Virtual Router Redundancy Protocol  
(VRRP) 343  
Rapid Spanning Tree Protocol (RSTP)  
Router Redundancy Protocol (RRP) snooping  
port security. See 802.1x Port-based Network Access  
Control; MAC address-based port security  
port trunks. See Link Aggregation Control Protocol (LACP)  
port trunk; static port trunks  
port-based access control. See 802.1x Port-based Network  
Access Control  
port-based VLANs  
routing interfaces  
private encryption key. See encryption key  
protected ports VLANs  
public encryption key. See encryption key  
Public Key Infrastructure (PKI)  
See also certificates, encryption keys  
certificates  
RRP snooping. See Router Redundancy Protocol (RRP)  
snooping  
S
Secure Shell (SSH)  
certification authority (CA)  
512  
Download from Www.Somanuals.com. All Manuals Search And Download.  
AT-S63 Management Software Features Guide  
source MAC addresses  
Secure Sockets Layer (SSL)  
See also certificates, encryption key  
server-based authentication. See RADIUS, TACACS+  
Simple Network Time Protocol (SNTP)  
Spanning Tree Protocol (STP)  
SSH. See Secure Shell (SSH)  
SSL. See Secure Sockets Layer (SSL)  
SNMP community strings  
static port trunks  
SNMPv1 and SNMPv2c  
T
TACACS+  
tagged VLANs  
SNMPv3 protocol  
513  
Download from Www.Somanuals.com. All Manuals Search And Download.  

3Com Crib Toy S24x User Manual
3M Snowshoes 747 User Manual
Acer Computer Monitor P203W User Manual
ADTRAN Network Card MX2800 M13 User Manual
ADT Security Services Home Security System 50075 User Manual
Agilent Technologies Sprinkler 54753a User Manual
Alliance Laundry Systems Clothes Dryer Drying Tumblers 50 Pound Capacity User Manual
Amana Range AGG222VDW User Manual
AMX Speaker System DAS KP LCDS W User Manual
Behringer Work Light UP1000 User Manual