H3C S9500E Series Routing Switches

IRF Configuration Guide

Hangzhou H3C Technologies Co., Ltd. http://www.h3c.com Software version: S9500E-CMW520-R1725 Document version: 6W170-20111118

Copyright 2011, Hangzhou H3C Technologies Co., Ltd. and its licensors

All rights reserved No part of this manual may be reproduced or transmitted in any form or by any means without prior written consent of Hangzhou H3C Technologies Co., Ltd. Trademarks , TOP G, , IRF, NetPilot, Neocean, NeoVTL, H3C, , Aolynk, , H3Care, SecPro, SecPoint, SecEngine, SecPath, Comware, Secware, Storware, NQA, VVG, V2G, VnG, PSPT, XGbus, N-Bus, TiGem, InnoVision and HUASAN are trademarks of Hangzhou H3C Technologies Co., Ltd. All other trademarks that may be mentioned in this manual are the property of their respective owners Notice The information in this document is subject to change without notice. Every effort has been made in the preparation of this document to ensure accuracy of the contents, but all statements, information, and recommendations in this document do not constitute the warranty of any kind, express or implied.

Preface
The H3C S9500E documentation set includes 14 configuration guides, which describe the software features for the H3C S9500E Series Routing Switches and guide you through the software configuration procedures. These configuration guides also provide configuration examples to help you apply software features to different network scenarios. The IRF Configuration Guide describes the Intelligent Resilient Framework (IRF) technology developed by H3C. It covers the basic concepts and working process of an IRF, and describes how to configure and access an IRF virtual device. This preface includes: Audience Conventions About the H3C S9500E documentation set Obtaining documentation Technical support Documentation feedback

Audience
This documentation is intended for: Network planners Field technical support and servicing engineers Network administrators working with the S9500E series

Conventions
This section describes the conventions used in this documentation set.

Command conventions
Convention
Boldface Italic [] { x | y | ... } [ x | y | ... ] { x | y | ... } *

Description
Bold text represents commands and keywords that you enter literally as shown. Italic text represents arguments that you replace with actual values. Square brackets enclose syntax choices (keywords or arguments) that are optional. Braces enclose a set of required syntax choices separated by vertical bars, from which you select one. Square brackets enclose a set of optional syntax choices separated by vertical bars, from which you select one or none. Asterisk marked braces enclose a set of required syntax choices separated by vertical bars, from which you select at least one.

Convention
[ x | y | ... ] * &<1-n> #

Description
Asterisk marked square brackets enclose optional syntax choices separated by vertical bars, from which you select one choice, multiple choices, or none. The argument or keyword and argument combination before the ampersand (&) sign can be entered 1 to n times. A line that starts with a pound (#) sign is comments.

GUI conventions
Convention
Boldface >

Description
Window names, button names, field names, and menu items are in Boldface. For example, the New User window appears; click OK. Multi-level menus are separated by angle brackets. For example, File > Create > Folder.

Symbols
Convention
WARNING CAUTION IMPORTANT NOTE TIP

Description
An alert that calls attention to important information that if not understood or followed can result in personal injury. An alert that calls attention to important information that if not understood or followed can result in data loss, data corruption, or damage to hardware or software. An alert that calls attention to essential information. An alert that contains additional or supplementary information. An alert that provides helpful information.

Network topology icons

Represents a generic network device, such as a router, switch, or firewall. Represents a routing-capable device, such as a router or Layer 3 switch. Represents a generic switch, such as a Layer 2 or Layer 3 switch, or a router that supports Layer 2 forwarding and other Layer 2 features.

Port numbering in examples

The port numbers in this document are for illustration only and might be unavailable on your switch.

About the H3C S9500E documentation set

The H3C S9500E documentation set includes: Category
Product description and

Documents
Marketing brochures

Purposes
Describe product specifications and benefits.

Category
specifications

Documents
Technology white papers Card datasheets Regulatory compliance and safety information Installation guide H3C N68 Cabinet Installation and Remodel Introduction H3C Pluggable SFP [SFP+][XFP] Transceiver Modules Installation Guide H3C High-End Network Products Hot-Swappable Module Manual Configuration guides

Purposes
Provide an in-depth description of software features and technologies. Describe card specifications, features, and standards. Provides regulatory information and the safety instructions that must be followed during installation. Provides a complete guide to hardware installation and hardware specifications. Guides you through installing and remodeling H3C N68 cabinets.

Hardware specifications and installation

Guides you through installing SFP/SFP+/XFP transceiver modules. Describes the hot-swappable modules available for the H3C high-end network products, their external views, and specifications. Describe software features and configuration procedures. Provide a quick reference to all available commands. Explains the system log messages. Explains the trap messages. Describes the MIBs for the software release. Provide information about the product release, including the version history, hardware and software compatibility matrix, version upgrade information, technical support information, and software upgrading.

Software configuration Command references System log messages Trap messages Operations and maintenance MIB Companion

Release notes

Obtaining documentation
You can access the most up-to-date H3C product documentation on the World Wide Web at http://www.h3c.com. Click the links on the top navigation bar to obtain different categories of product documentation: [Technical Support & Documents > Technical Documents] Provides hardware installation, software upgrading, and software feature configuration and maintenance documentation. [Products & Solutions] Provides information about products and technologies, as well as solutions. [Technical Support & Documents > Software Download] Provides the documentation released with the software version.

Technical support
[email protected] http://www.h3c.com

Documentation feedback
You can e-mail your comments about product documentation to [email protected]. We appreciate your comments.

Contents
Configuring IRF 1 IRF overview 1 Benefits 1 Application scenario 1 Basic concepts 2 Establishment, operation, and maintenance of an IRF fabric 4 Connecting the IRF member switches 5 Topology collection 6 Master election 7 IRF fabric management and maintenance 7 IRF fabric configuration task list 8 Configuration prerequisites 9 Pre-configuration 9 Non pre-configuration 10 Pre-configuring an IRF member switch in standalone mode 11 Binding a physical port to an IRF port 12 Setting a member ID for the switch 13 Specifying a priority for the switch 13 Enabling enhanced IRF mode 13 Configuration guidelines 13 Configuration procedure 14 Saving the running configuration to the configuration file to be used at the next system startup 14 Switching operating mode 14 IRF modes 14 Configuration file auto-conversion 15 Switching the operating mode to IRF mode 15 Accessing an IRF fabric 15 Accessing the active MPU of an IRF fabric 15 Accessing a standby MPU of an IRF fabric 15 Configuring IRF member switches in IRF mode 16 Specifying an IRF domain ID for an IRF fabric 16 Configuring IRF ports 17 Setting a member ID for a switch 19 Specifying a priority for a member switch 19 Configuring a description for a member switch 19 Specifying the preservation time of the bridge MAC address 20 Enabling auto reboot for IRF fabric merge 21 Enabling automatic boot file updating 21 Setting the IRF link down report delay 22 Enabling IRF link failure detection and auto-recovery 23 Configuring MAD detection 23 Performing IRF configuration fast recovery 33 Configuration prerequisites 33 Configuration procedures 33 Displaying and maintaining an IRF fabric 34 IRF fabric configuration examples 34 LACP MAD detection-enabled configuration example (non pre-configuration mode) 35 BFD MAD detection-enabled IRF configuration example (pre-configuration mode) 38 ARP MAD detection-enabled IRF configuration example (pre-configuration mode) 40
i

Switching the operating mode of IRF member switches from IRF to standalone 43 Enhanced IRF mode configuration example (four switches forming an IRF fabric) 45

Index 1

ii

Configuring IRF
IRF overview
You can use the H3C Intelligent Resilient Framework (IRF) technology to connect and virtualize multiple switches into a virtual switch called an IRF fabric to provide data center class availability and scalability. IRF virtualization technology offers processing power, interaction, unified management, and uninterrupted maintenance of multiple switches. NOTE: Up to four switches of the same series can form an IRF fabric.

Benefits
IRF delivers the following benefits: Simplified topology and streamlined management. An IRF fabric appears as one node on the network. You can log in at any member switch to manage all members of the IRF fabric. Network scalability and resiliency. You can increase ports and network bandwidth of an IRF fabric simply by adding member switches. High availability and reliability. The member switches in an IRF fabric work in redundant mode. One member switch works as the master to manage and maintain the entire IRF fabric, and other member switches process services and provide backup. If the master fails, another member switch is elected as the new master to prevent service interruption. You can perform link aggregation not only for IRF links but also for physical links between the IRF fabric and its upper or lower layer devices for link redundancy.

Application scenario
Figure 1 shows an IRF fabric that comprises two switches, which appear as a single node to the upper and lower layer switches.

Figure 1 IRF application scenario

Basic concepts
Figure 2 IRF implementation schematic diagram
Device A (MemberID=1)
Active MPU of the member Standby MPU of the member

Device B (MemberID=2)
Active MPU of the member Standby MPU of the member

IRF-Port2

IRF-Port1

Service XGE1/3/0/1 interface Physical IRF port

IRF link

XGE2/3/0/1 Service Physical IRF interface port

After an IRF is formed. Suppose Device A is the master.

IRF

Master (MemberID=1)
Active MPU of the IRF Standby MPU of the IRF

Slave (MemberID=2)
Standby MPU of the IRF Standby MPU of the IRF

Device A and Device B in Figure 2 form an IRF fabric which has four MPUs (one active and three standby) and two interface cards. The IRF fabric manages both the physical and software resources of Device A and Device B. This section uses Figure 2 to explain the concepts that you might encounter when working with IRF.

Operating mode
A switch can operate in either of the following two modes: Standalone modeThe switch cannot form an IRF fabric with other switches. IRF modeThe switch can connect with other switches to form an IRF fabric.

You can change the operating mode of a switch at the command line interface (CLI).

IRF member switch roles

IRF uses two member switch roles: master and slave. When switches form an IRF fabric, they elect a master to manage the IRF fabric, and the other switches back up the master. When the master switch fails, another switch is elect as the new master to avoid service interruption. For more information about master election, see Master election.

Roles of MPUs
Each member switch in an IRF fabric has one or two MPUs (also called the supervision engines) and they play different roles, as follows: Role
Local active main processing unit (MPU) Local standby MPU

Description
The supervisor engine that manages the local switch. The supervisor engine backs up the local active MPU, and takes over when the local active MPU fails. The active MPU of the master switch. You configure and manage the entire IRF at the command line of the global active MPU. All MPUs except the active MPU of the master switch are global standby MPUs.

Global active MPU

Global standby MPU

IRF member ID
You assign the active MPU of each member switch a unique ID to identify the switch in the IRF fabric. This ID is called the IRF member ID of the switch. By default, the standby MPU of a switch is automatically assigned the same ID as the active MPU. You can change the standby MPU ID of a member switch to quickly recover IRF configuration for a switch that has only one MPU as described in Performing IRF configuration fast recovery.

IRF port
An IRF port is a logical interface for the internal connection between IRF member switches. Each IRF member switch has two IRF ports: IRF-port 1 and IRF-port 2. An IRF port is activated when you bind a physical port to it. NOTE: In standalone mode, the IRF ports are named IRF-port1 and IRF-port2. In IRF mode, the IRF ports are named IRF-portn/1 and IRF-portn/2, where n is the member ID of the switch. In this manual, IRF-port1 and IRF-port2 are used.
3

Physical IRF port

Physical IRF ports are physical ports bound to an IRF port. They connect IRF member switches and forward IRF protocol packets and data packets between IRF member switches. Physical IRF ports can be electrical ports or optical ports.

IRF partition
IRF partition occurs when an IRF fabric splits into two or more IRF fabrics because of IRF link failures, as shown in Figure 3. The partitioned IRF fabrics operate with the same IP address and cause routing and forwarding problems on the network. Figure 3 IRF partition

NOTE: When you unplug the card where a physical IRF port resides, IRF partition might occur. Therefore, when you unplug this kind of card, make sure that there are at least two physical IRF ports in UP state, and they are not on the same card.

IRF merge
IRF merge occurs when two partitioned IRF fabrics re-unite or when you configure and connect two independent IRF fabrics to be one IRF fabric, as shown in Figure 4. Figure 4 IRF merge
IRF 1 IRF 2 IRF

+
Device A Device B

=
Device A

XGE1/3/0/1

XGE2/3/0/1

IRF link Device B

Member priority
Member priority determines the role of a member switch during the master election process. A member with a higher priority is more likely to be a master. The priority of a switch defaults to 1. You can modify the priority at the CLI.

Establishment, operation, and maintenance of an IRF fabric

IRF fabric management involves these stages: Connecting the IRF member switches, Topology collection, Master election, and IRF fabric management and maintenance.

Connecting the IRF member switches

Connection medium
To establish an IRF fabric, physically connect the physical IRF ports of member switches. The connection medium depends on the physical IRF ports supported by the switch. If you use electrical interfaces as physical IRF ports, use network cables (cross-over or straight-through) to connect them. This connection mode improves the usage of the available resources (electrical interfaces are used to forward data traffic when not bound to any IRF port, and used to forward packets between member switches when bound to IRF ports), and saves the cost as well (without the need to purchase optical module used for IRF connection). If you use optical ports as physical IRF ports, use fibers to connect them. This connection mode connects physical switches located very far at a distance and provides flexible application.

NOTE: A good practice is to use 10G optical Ethernet interfaces as physical IRF ports.

Connecting requirements
As shown in Figure 5, connect the physical ports bound to IRF-Port1 on one switch to the physical ports bound to the IRF-Port2 on its neighbor switch. Figure 5 IRF fabric physical connection

NOTE: An IRF port can be bound to a maximum of 12 physical ports to increase the bandwidth and reliability of the IRF port.

IRF topologies
IRF member switches typically adopt a daisy chain topology or ring topology, as shown in Figure 6. The daisy chain topology is mainly used in a network where member switches are distributedly located. The ring topology is more reliable than the daisy chain topology. In a daisy chained IRF fabric, the failure of one link can cause the IRF fabric to partition into two independent IRF fabrics; the failure of a link in a ring topology result in a daisy chain connection, not affecting IRF services.

If two IRF member switches are far away from each other (for example, if they are in different cities), you can use a relay device to connect them, as shown in Figure 7.

Figure 6 IRF connections

Figure 7 IRF connections

Topology collection
Each member exchanges IRF hello packets with neighbors to collect the topology data, including IRF port connection states, member IDs, priorities, and bridge MAC addresses. Each member is managed by its active MPU, which records its known topology information locally. At the startup of a member switch, the active MPU of the member switch records topology information of the member switch. When an IRF port of the member switch is up, the active MPU of the switch performs the following operations:
1. 2. 3.

Periodically sends its known topology information from this port. When receiving the topology information from the directly connected neighbor, it updates the local topology information. If a standby MPU is available on the member switch, the active MPU synchronizes its recorded topology information to the standby MPU to ensure consistent topology information on both boards.
6

After all member switches have obtained topology information (known as topology convergence), the IRF fabric enters the role election stage.

Master election
Master election is held each time the topology changes, for example, when the IRF fabric is established, a new member switch is plugged in, the master switch fails or is removed, or the partitioned IRF fabrics merge. The master is elected based on the following rules in descending order: The current master, even if a new member has a higher priority. (When an IRF fabric is being formed, all member switches consider themselves as the master, and this rule is skipped) The switch with a higher priority. The switch with the longest system up-time. (The member switches exchange system up-time in the IRF hello packets) The switch with the lowest bridge MAC address.

The IRF fabric is formed on election of the master. NOTE: During an IRF merge, an IRF election is held, the switches of the IRF fabric that fails the master election must reboot to re-join the IRF fabric that wins the election. Then, the switch reboots with the execution of a command. After a master election, all slave member switches initialize and reboot with the configuration on the master, and their original configuration, even if has been saved, will be lost.

IRF fabric management and maintenance

After the IRF fabric is established, you can access the master from any member switch to manage all the resources of the member switches.

Member ID
An IRF fabric uses member IDs to uniquely identify and manage its members. For example, if an interface on a switch that operates in standalone mode is named GigabitEthernet 3/0/1. After the switch joins an IRF fabric, it receives a member ID of 2. The name of the interface changes to GigabitEthernet 2/3/0/1. Member ID is also used in file management. For example, when the switch operates in standalone mode, the path of a file was slot1#flash:/test.cfg. After the switch joins an IRF fabric, the path changes to chassis1#slot1#flash:/test.cfg, which indicates that the file is saved on the board in slot 1 of member switch 1. Therefore, member IDs must be unique. NOTE: Member IDs and priorities are configured per switch. If you set the member ID or priority for a member switch, the configuration is first saved on the active MPU of the member switch, and then synchronized to the standby MPU. If the active MPU and standby MPU of a member switch keep different member IDs, the member ID kept by the active MPU is applied. For example, if the switch with the member ID of 2 has only one active MPU, after you plug in a standby MPU that keeps a member ID of 1, the member ID of the switch is still 2 and the member ID kept on the standby MPU is synchronized to 2.

IRF fabric topology maintenance

As soon as a member switch is down or an IRF link is down, its neighbor switches broadcast the leaving of the switch to other members. When a member switch receives the leave message, it looks up its IRF topology database to determine whether the leaving switch is the master. If yes, the member switch starts a master election and updates its IRF topology database. If the leaving switch is not a master, the member switch directly updates its IRF topology database. NOTE: An IRF port goes down only when all its physical IRF ports are down.

IRF multi-active detection

An IRF link failure causes an IRF fabric to split in two IRF fabrics operating with the same Layer 3 configurations, such as the same IP address. To avoid IP address collisions and network problems, IRF uses the multi-active detection (MAD) mechanism to detect the presence of multiple identical IRF fabrics and handle collisions. MAD provides the following functions:
1.

Detection

MAD detects multiple identical active IRF devices with the same global configuration by extending the Link Aggregation Control Protocol (LACP), the Bidirectional Forwarding Detection (BFD) protocol, or the Gratuitous Address Resolution (ARP) protocol. For more information, see Configuring MAD detection.
2.

Collision handling

If multiple identical active IRF fabrics are detected, only the IRF fabric that has the lowest master ID can operate in active state and forward traffic normally. MAD sets all other IRF fabrics in recovery state (disabled) and shuts down all physical ports but the console and physical IRF ports and other ports you have specified with the mad exclude interface command.
3.

Failure recovery

An IRF link failure triggers IRF fabric partition and causes multi-active collision. In this case, repair the failed IRF link to make the collided IRF fabrics merge into one and recover the failure. If the IRF fabric in recovery state fails before the failure is recovered, repair both the failed IRF fabric and the failed IRF link, and then the collided IRF fabrics can merge into one and the failure is recovered. If the IRF fabric in active state fails before the failure is recovered, enable the IRF fabric in recovery state at the CLI to make it take over the active IRF fabric and protect the services from being affected. Then, recover the MAD failure. NOTE: For information about LACP, see Layer 2LAN Switching Configuration Guide. For information about BFD, see High Availability Configuration Guide. For information about gratuitous ARP, see Layer 3IP Services Configuration Guide.

IRF fabric configuration task list

There are two IRF fabric configuration methods, pre-configuration and non pre-configuration. The pre-configuration method is recommended because only one switch reboot is needed to complete the configuration.

Configuration prerequisites
Configure a higher priority for a member switch that features good performance and rich functions so that the switch can be elected as the master when two switches form an IRF fabric for the first time. If you use a physical port of a switch as a physical IRF port when the switch operates in standalone mode, the services configured on the port will become ineffective after the switch switches to IRF mode. Make preparations in advance to ensure services are not affected. Before establishing an IRF fabric, make sure that the system working mode of the member switches is the same. If not, the IRF fabric cannot be established. For more information about the system working mode, see Fundamentals Configuration Guide. The member switches of an IRF fabric must work in the same rule match mode. This means that you must configure the acl ipv6 enable command, or the acl ipv6 disable command on the switches. For more information about the acl ipv6 command, see ACL and QoS Configuration Guide. The member switches of an IRF fabric must be configured with the same VPN label processing mode. This means that you must configure the vpn popgo command or the undo vpn popgo command on the switches. For more information about the vpn popgo command, see ACL and MPLS Configuration Guide. Before establishing an IRF fabric, check that enhanced IRF mode is enabled on all member switches or disabled on all member switches. If enhanced IRF mode is enabled on some member switches but disabled on the others, the IRF fabric cannot be established. For more information about enhanced IRF mode, see Enabling enhanced IRF mode.

Pre-configuration
Pre-configuration allows you to configure IRF ports, member IDs, and member priority on a switch operating in standalone mode. These configurations do not affect the running of the switch, and take effect only when the switch switches to IRF mode. Adopt this method to configure an IRF fabric before establishing it. To enable switches operating in standalone mode to form an IRF fabric, you must pre-configure their member IDs and switch their operating modes. If you configure the priority for a switch as the greatest value when the switch operates in standalone mode, this switch can win the master election and become the master after multiple switches form an IRF fabric for the first time. If you configure IRF ports for member switches operating in standalone mode, they can directly form an IRF fabric with other switches after their operating mode is switched to IRF. Pre-configurations enable the member switches to reboot only once to establish an IRF fabric. Complete these tasks to configure an IRF fabric in pre-configuration mode: Task
Pre-configuring an IRF member switch in standalone mode Binding a physical port to an IRF port Setting a member ID for the switch Specifying a priority for the switch Enabling enhanced IRF mode Saving the running configuration to the configuration file to be used at the next system startup

Remarks
Required Required Optional Optional Perform this task to use three or four switches to form an IRF fabric. Required

Task

Remarks
Required Optical Ethernet interfaces and electrical Ethernet interfaces (except combo ports) can be used as physical IRF ports. Be sure to connect the switches according to the requirements described in the section Connecting requirements Otherwise, they cannot form an IRF fabric. Required Required Required Optional Optional Optional Optional Optional Optional Optional Optional Optional

Physically connecting standalone mode

two

switches

operating

in

Switching operating mode Activating the physical IRF ports of the member switches by using the undo shutdown command Accessing an IRF fabric Accessing the active MPU of an IRF fabric Accessing a standby MPU of an IRF fabric Configuring a description for a member switch Specifying the preservation time of the bridge MAC address Enabling auto reboot for IRF fabric merge Configuring IRF member switches in IRF mode Enabling updating automatic boot file

Setting the IRF link down report delay Enabling IRF link failure detection and auto-recovery Configuring MAD detection Performing IRF configuration fast recovery

Non pre-configuration
Non pre-configuration allows you to configure the member ID for a switch operating in standalone mode, switch the operating mode to IRF mode, and then configure parameters such as a new member ID and member priority (during the whole process, the member switches may reboot for multiple times). Use this method when you need to modify the running configuration. For example, Change the member ID of a switch to a specified value.

NOTE: Changing member ID might cause ineffectiveness of some member ID-related configurations. Modify the priority of a member switch to make sure it is elected as the master in the next master election.
10

Modify the binding between an IRF port and physical IRF ports (such as deleting a binding or adding a new binding), and the configuration of IRF ports may affect the operation of the switch (for example, causing IRF partition, or IRF merge). Task
Setting a member ID for the switch Enabling enhanced IRF mode Saving the running configuration to the configuration file to be used at the next system startup Switching operating mode Accessing fabric an IRF Accessing the active MPU of an IRF fabric Accessing a standby MPU of an IRF fabric Configuring IRF ports Setting a member ID for the switch Specifying a priority for a member switch

Complete these tasks to configure an IRF fabric in non pre-configuration mode: Remarks
Required Optional Perform this task to use three or four switches to form an IRF fabric. Required Required Required Optional Required Optional Optional Required Physically connecting two operating in standalone mode switches Be sure to connect the switches according to the requirements described in the section Connecting requirements. Otherwise, they cannot form an IRF fabric. Required Optional Optional Optional Optional Optional Optional Optional

Configuring IRF member switches in IRF mode

Configuring a description for a member switch Specifying the preservation time of the bridge MAC address Enabling auto reboot for IRF fabric merge Enabling automatic boot file updating Setting the IRF link down report delay Enabling IRF link failure detection and auto-recovery Configuring MAD detection Performing IRF configuration fast recovery

Pre-configuring an IRF member switch in standalone mode

You can configure the IRF ports, member ID, and member priority for the switch when it is operating in standalone mode. Configurations take effect when the operating mode of the switch switches to IRF.

11

Binding a physical port to an IRF port

To establish IRF connection, you must assign the physical ports that connect IRF member switches to IRF ports. An IRF port can be bound to a maximum of 12 physical ports and is known as an aggregate IRF port, which can be realized by repeatedly executing the port group interface command. This allows two switches to be connected through 12 Ethernet cables or fibers to increase the bandwidth and reliability of the IRF port. To bind a physical port to an IRF port: To do
Enter system view Enter physical IRF port view

Use the command

system-view interface interface-number interface-type

Remarks
Required when two member switches have more than one pair of physical IRF ports connected. Required

Shut down the port

shutdown

Return to system view Create an IRF port and enter IRF port view when the switch operates in standalone mode (if the IRF port is already created, this command enters IRF port view) Bind a physical IRF port to the IRF port

quit

irf-port port-number

By default, no IRF port is created on the switch. Required By default, an IRF port is not bound to any physical IRF port. Optional Make sure that the binding is as expected. If the binding is incorrect, IRF cabling errors may occur, resulting in IRF establishment failure.

port group interface interface-type interface-number

Verify the binding configuration

display irf configuration [ | { begin | exclude | include } regular-expression ]

CAUTION: A combo port cannot be bound to an IRF port. For information about combo ports, see Interface

Configuration Guide.

Save the configurations to the startup configuration file so that the configurations can take effect when the switch is switched to IRF mode. In standalone mode, binding a physical port to an IRF port does not affect the running configuration of the port. However, when the operating mode changes to IRF mode, the default configuration of the physical IRF port restores, and you can only execute the shutdown and description commands on the physical port. For more information about the shutdown and description commands, see Interface Command Reference.

12

Setting a member ID for the switch

A switch by default operates in standalone mode without an IRF member ID. You must assign it an IRF member ID before you can set it in IRF mode. You can execute the display irf configuration command and check the MemberID field. If the switch has no IRF member ID, the field displays two hyphens (--). To avoid member ID collision with other members when the switch is added into an IRF fabric, plan the member ID of this switch when it operates in standalone mode. To do
Enter system view Set a member ID for the switch when it operates in standalone mode

To set a member ID for the switch: Use the command

system-view irf member member-id

Remarks
Required By default, no member ID is set for the switch.

Specifying a priority for the switch

To specify a priority for the switch: To do
Enter system view Specify a priority for the switch when it operates in standalone mode

Use the command

system-view irf priority priority

Remarks
Optional The priority of the switch defaults to 1.

Enabling enhanced IRF mode

The enhanced IRF mode allows you to create an IRF fabric comprising up to four member switches. Both standalone and IRF modes support enhanced IRF mode.

Configuration guidelines
Follow these guidelines: In standalone mode, you can enable enhanced IRF mode directly. In IRF mode,

If no Layer-3 Ethernet ports exist and no VPLS and MAC-in-MAC instances exist, you can enable enhanced IRF mode directly. If Layer-3 Ethernet ports exist, you must change them to Layer-2 Ethernet ports (prompt information available) before you can enable enhanced IRF mode. If no Layer-3 Ethernet ports exist but VPLS or MAC-in-MAC instances exist, you must reboot the switch (prompt information available) before you can enable enhanced IRF mode.

For more information about Layer-3 Ethernet ports, see Interface Configuration Guide. For more information about VPLS, see MPLS Configuration Guide. For more information about MAC-in-MAC, see Layer-2----LAN Switching Configuration Guide.
13

You must use the save command to save the configuration after you perform this task. To disable enhanced IRF mode with the undo irf mode enhanced command after you enable enhanced IRF mode in IRF mode, the member switches must be no more than two and each member switch must have only one IRF port; otherwise, the enhanced IRF mode cannot be disabled. Before emerging IRF fabrics, check that enhanced IRF mode is enabled on all member switches or disabled on all member switches. If enhanced IRF mode is enabled on some member switches but disabled on the others, the IRF fabrics cannot be merged. Before IRF merge, if enhanced IRF mode is enabled on all IRF member switches, you must reboot all IRF members except any one of them to complete IRF merge.

Configuration procedure
To enable enhanced IRF mode: To do
Enter system view Enable enhanced IRF mode

Use the command

system-view irf mode enhanced

Remarks
Required Not enabled by default.

Saving the running configuration to the configuration file to be used at the next system startup
To save the running configuration to the configuration file to be used at the next system startup: To do
Save the running configuration to the configuration file to be used at the next system startup

Use the command

save [ safely ] [ force ]

Remarks
Required Available in any view.

Switching operating mode

IRF modes
The switch works in either standalone or IRF mode. A standalone switch does not belong to any IRF fabric. To assign a standalone switch to an IRF fabric, you must switch the operating mode to IRF mode before adding it to an IRF fabric. An IRF mode switch can itself form an IRF fabric or form an IRF fabric with other IRF mode switches. To save management costs and system resources, set a switch in IRF mode only when you are creating a multi-member IRF fabric.

By default, a switch is operating in standalone mode. The switch reboots when its operating mode changes to IRF mode.

14

Configuration file auto-conversion

When you change the operating mode of the switch from standalone to IRF, you can use the configuration file auto-conversion function to convert the startup configuration file to prevent some slot- or interface-related configurations from becoming invalid. For example, this function can convert the slot slot-number parameter set in standalone mode to the chassis chassis-number slot slot-number parameter in IRF mode, and add the chassis ID in an interface number.

Switching the operating mode to IRF mode

To switch the operating mode of the switch to IRF mode: To do
Enter system view Switch the operating mode of the switch to IRF mode

Use the command

system-view chassis convert mode irf

Remarks
Required The default operating mode is standalone mode.

NOTE: To display the member ID of the switch, use the display irf configuration command. If the MemberID field is displayed as , the member ID of the switch is not set yet, and you need to use the irf member command to configure it. When you switch the operating mode, the switch reboots automatically to make the change effective. H3C recommends that you save the running configuration before switching the operating mode. To switch the operating mode of the switch to standalone mode, use the undo chassis convert mode irf command.

Accessing an IRF fabric

Accessing the active MPU of an IRF fabric
Access an IRF fabric in one of the following ways: Local loginLog in on the AUX or console port of a member switch. Remote loginRemotely log in at a Layer 3 Ethernet interface on any member switch through Telnet, the web, or SNMP.

When you log in to an IRF fabric, you are placed at the CLI of its active MPU, regardless of at which member switch you are logged in. The active MPU of the IRF fabric is the configuration and control center of the IRF fabric. You make configuration for the IRF fabric on the active MPU, and the IRF fabric synchronizes the configurations to all standby MPUs in the virtual IRF device.

Accessing a standby MPU of an IRF fabric

You can log in to the CLI of a standby MPU of the IRF fabric to display its configurations and debug the standby MPU. When you switch from the active MPUs CLI to the standby MPUs CLI, you are placed in the user view of the standby MPU and the command prompt changes to

15

<Sysname-Slave#member-ID/slot-number>, for example, <Sysname-Slave#1/0>. You can perform the following commands at the CLI of the standby MPU of an IRF fabric: display quit return system-view debugging terminal debugging terminal logging terminal monitor terminal trapping

To return to the CLI of the active MPU of the IRF fabric, use the quit command. To log in to the CLI of a standby MPU in the IRF fabric: To do Use the command Remarks
Required Log in to a standby MPU of an IRF fabric irf switch-to chassis chassis-number slot slot-number By default, you are placed at the CLI of the active MPU of the IRF fabric when you log in to the IRF fabric. Available in user view

Configuring IRF member switches in IRF mode

Specifying an IRF domain ID for an IRF fabric
Introduction to domain
IRF uses IRF domain IDs to uniquely identify IRF fabrics. IRF domain IDs prevent IRF fabrics from interfering with one another. See Figure 8, Device A and Device B form IRF fabric 1, and Switch A and Switch B form IRF fabric 2. The fabrics have LACP MAD detection links between them. When a member switch in one IRF fabric receives an extended LACP packet for MAD detection, it looks at the domain ID in the packet to see whether the packet is from the local IRF fabric or from a different IRF fabric. Then, the switch can handle the packet correctly.

16

Figure 8 A network that comprises two IRF domains

Core network

Device A

IRF 1 (domain 10)

XGE1/3/0/1

Device B

IRF link

XGE2/3/0/1

Switch A

IRF 2 (domain 20)

IRF link

Switch B

Access network

Assigning a domain ID to an IRF fabric

To assign a domain ID to an IRF fabric: To do
Enter system view

Use the command

system-view

Remarks
Required if LACP MAD or ARP MAD is adopted.

Assign a domain ID to the IRF fabric

irf domain domain-id

Optional adopted.

if

BFD

MAD

is

By default, the domain ID of an IRF fabric is 0.

Configuring IRF ports

To bring the IRF function into work, you must connect the IRF member switches, assign the connecting physical ports to the appropriate IRF port on each member switch, and activate the IRF port configuration. After the IRF port configuration is activated, the IRF ports go up, a master election is held, and the switches that has failed in the election automatically reboot to join the IRF fabric as slave switches. To configure IRF ports: To do
Enter system view

Use the command

system-view

Remarks

17

To do
Enter physical IRF port view Shut down the port Return to system view Create an IRF port and enter IRF port view Bind a physical IRF port to the IRF port

Use the command

interface interface-number shutdown quit irf-port member-id/port-number interface-type

Remarks
Required Required By default, no IRF port is created on the switch. Required By default, an IRF port is not bound to any physical IRF port. Optional Make sure that the binding is as expected. If the binding is incorrect, IRF cabling errors may occur, resulting in IRF establishment failure. interface-type Required

port group interface interface-type interface-number

Verify the binding configuration

display irf configuration [ | { begin | exclude | include } regular-expression ]

Return to system view Enter physical IRF port view Bring up the port Return to system view

quit interface interface-number undo shutdown quit

NOTE: After you bind a physical port to an IRF port, all services configured before the binding on the physical port become invalid. Make sure that the previous services are not interrupted. To bind several physical IRF ports to an IRF port, execute the port group interface command multiple times. Aggregate IRF ports expand network bandwidth and provide link redundancy. The maximum number of physical IRF ports that can be bound to an IRF port is 12. When the upper limit is reached, the execution of this command fails. Before you create or remove an IRF port binding, always shut down the physical IRF port. After you are finished, perform the undo shutdown command to bring up the port. If the system prevents you from shutting down a port for avoiding anomalies, follow the system instructions to disable its peer port. A combo port cannot be bound to an IRF port. For information about combo ports, see Interface

Configuration Guide.

If a common Ethernet interface functions as a physical IRF port and is bound to an IRF port, you can execute the shutdown, description, flow-control, oam enable, oam loopback, and oam mode commands only on the physical IRF port. For more information about the shutdown, description, and flow-control commands, see Interface Command Reference. To make sure that the configuration takes effect at the next switch startup, save configurations to the startup configuration file that will be used at the next startup after binding a physical port to an IRF port or canceling the binding.

18

Setting a member ID for a switch

An IRF fabric uses member IDs to uniquely identify its members. A lot of information and configurations relate to member IDs, such as port (physical or logical) numbers, configurations on ports, and member priorities. After you change the member ID of a switch: If you do not reboot the switch, the original member ID still takes effect and all physical resources are identified by the original member ID. In the configuration file, only the IRF port numbers, configurations on IRF ports, and priority of the switch change with the member ID, other configurations do not change. If you save the running configuration and reboot the switch, the new member ID takes effect and all physical resources are identified by the new member ID. In the configuration file, only the IRF port numbers, configurations on IRF ports, and priority of the switch still take effect, other configurations (such as configuration for physical IRF ports, value of the chassis keyword which adopts the original member ID) no longer take effect and you need to configure them again. To do
Enter system view Set a member ID for a member switch

To set a member ID for a switch: Use the command

system-view irf member member-id new-member-id renumber

Remarks
Optional The member ID of a switch defaults to 1.

CAUTION: Member ID changes take effect at the reboot of the switch. Change member IDs for the switches in an IRF fabric with caution. The change might cause configuration change and even data loss.

Specifying a priority for a member switch

To specify a priority for a member switch: To do
Enter system view Specify a priority for a member switch

Use the command

system-view irf member member-id priority priority

Remarks
Optional The priority of a member defaults to 1.

Configuring a description for a member switch

You can configure a description for a member switch to identify its physical location, or for any other management purpose. To configure a description for a member switch:

19

To do
Enter system view Configure the description for a member switch

Use the command

system-view irf member member-id description text

Remarks
Optional Not configured by default.

Specifying the preservation time of the bridge MAC address

A switch uses the bridge MAC address when it communicates with the external networks as a bridge. Some Layer 2 protocols (like LACP) use bridge MAC addresses to identify different switches. During Layer 2 packet forwarding, if the destination MAC address of a packet is the bridge MAC address of a switch, the packet is sent to this switch; otherwise, the packet is discarded. This means the bridge switch on your network must have a unique bridge MAC address. If multiple devices on your network have the same bridge MAC addresses, bridge MAC address collision occurs and the communication fails. An IRF fabric communicates with external networks as a single switch and has a bridge MAC address. Typically, an IRF fabric uses the bridge MAC address of its master. Bridge MAC address collision causes communication failure, and bridge MAC address switching causes traffic interruption. Properly configure the preservation time of the bridge MAC address of the IRF fabric to avoid this issue. Bridge MAC address collision occurs and causes network communication problems when a master leaves an IRF fabric to join another IRF fabric or to operate in a standalone manner and the IRF fabric is configured to permanently preserve the bridge MAC address, bridge MAC address collision occurs and thus causes network communication problem. Unnecessary bridge MAC address switching occurs causing the flow interruption when the master leaves the IRF fabric because of a reboot or link failure, and the IRF fabric is configured to change its bridge MAC address as soon as the master leaves.

To avoid these issues, configure the preservation time of the bridge MAC address of the IRF fabric according to your network status: Preserve for six minutes: When the master leaves, the bridge MAC address does not change for six minutes. If the master does not come back within six minutes, the IRF fabric uses the bridge MAC address of the newly elected master as its bridge MAC address. If the master leaves for a short time because of a master reboot or link failure, but then returns to the IRF fabric, specify this setting to minimize flow interruption caused by unnecessary bridge MAC address switch. Preserve permanently: The bridge MAC address of the IRF fabric does not change no matter how long the master leaves the IRF fabric. Not preserved: When the master leaves, the IRF fabric uses the bridge MAC address of the newly elected master as its bridge MAC address. To do
Enter system view Configure the bridge MAC address of the IRF fabric to be preserved permanently when the master leaves

To specify the preservation time of the bridge MAC address of the IRF fabric: Use the command
system-view

Remarks
Optional

irf mac-address persistent always

By default, the bridge MAC address of the IRF fabric is

20

To do
Specify the preservation time of the bridge MAC address of the IRF fabric as 6 minutes when the master leaves Configure that the bridge MAC address of the IRF fabric changes as soon as the master leaves

Use the command

Remarks
preserved permanently when the master leaves.

irf mac-address persistent timer

undo irf mac-address persistent

CAUTION: Bridge MAC address change may cause a temporary traffic interruption. If two IRF fabrics have the same bridge MAC address, they cannot be merged into one IRF fabric. If you use ARP MAD together with the spanning tree feature for an IRF fabric, enable the IRF fabric to change its bridge MAC address as soon as the master leaves by using the undo irf mac-address persistent command. If the VRRP load balancing function is enabled when the switch operates in IRF mode, you must configure the preservation time of IRF bridge MAC address as permanently preserved (which is the default setting). For more information about VRRP, see High Availability Configuration Guide.

Enabling auto reboot for IRF fabric merge

When merging IRF fabrics, you must reboot the member switches in the IRF fabric that has been defeated in the master election. The auto reboot function enables the IRF fabric to automatically reboot all its member switches to complete the merge. This function can work only when it is enabled on both IRF fabrics that are merging. To enable auto reboot for IRF fabric merge: To do
Enter system view Enable auto reboot for IRF fabric merge

Use the command

system-view irf auto-merge enable

Remarks
Optional Disabled by default.

NOTE: The auto reboot function does not take effect on the IRF fabric merge caused by binding a physical port to an IRF port in IRF mode. You must manually reboot the switches that have been defeated in the master selection to complete the merge.

Enabling automatic boot file updating

When auto upgrade of boot files is disabled and the boot files of slaves and the master are different versions, the new member or the member with a low priority do not boot normally. Update the switch version and add the switch into the IRF fabric again. When auto upgrade of boot file is enabled, the IRF fabric compares its software version with that of the master as soon as a switch is added into an IRF fabric. If the versions are not the same, the switch automatically downloads the boot file from the master, reboots with the new boot file, and
21

joins the IRF fabric again. If the downloaded boot file and the local boot file have duplicate file names, the local file is overwritten. To enable auto upgrade of boot files for an IRF fabric: To do
Enter system view Enable auto upgrade of boot files for an IRF fabric

Use the command

system-view irf auto-update enable

Remarks
Optional Enabled by default

CAUTION: When you disable the auto upgrade of boot files, the software versions of the switches must be the same. If not, an IRF fabric cannot be established. Make sure that the model of the switch matches the software version of the master before adding the switch into the IRF fabric. If the model of the switch to be added does not match the software version of the master, auto upgrade may not function normally. After automatically loading the masters boot file, a slave configures the file as the boot file to be used at the next boot and reboots automatically.

Make sure that there is enough space on the slaves flash or CF card where the system boot file resides to make the auto upgrade succeeds.

Setting the IRF link down report delay

After setting the delay time for the link layer to report a link-down event: If the IRF link state changes from up to down, the port does not immediately report the change to the IRF fabric. If the IRF link state is still down when the configured time is reached, the port reports the change to the IRF fabric. If the link state changes from down to up, the link layer immediately reports the event to the IRF fabric.

Use this function to avoid unnecessary IRF fabric partition and merge caused by frequent link state changes of a port. To set the IRF link down report delay: To do
Enter system view Set the IRF link down report delay

Use the command

system-view irf link-delay interval

Remarks
Optional 0 milliseconds by default.

CAUTION: The recommended value of the interval argument is in the range 200 to 500, in milliseconds. Setting the delay time too long can cause the IRF fabric to not address topology changes in time, which can slow the recovery process.

22

Enabling IRF link failure detection and auto-recovery

IRF link failure detection and auto-recovery automatically check the health of IRF links and automatically make link recovery attempts when a failed IRF link is detected. The two functions are helpful for IRF fabrics that have more than one IRF link. To enable IRF link failure detection and auto-recovery: To do
Enter system view

Use the command

system-view

Remarks
Optional Enabled by default.

Enable IRF link failure detection

irf link-status detect enable

IRF link failure detection must be enabled when a relay device is present between two member devices. Optional Enabled by default.

Enable IRF link auto-recovery

irf link-status auto-recovery enable

You can enable the IRF link auto-recovery function only when the IRF link failure detection function is enabled.

When the IRF link failure detection function detects an IRF link failure, it handles the physical IRF port depending on the status of IRF link auto-recovery: If IRF link auto-recovery is enabled, disables the physical IRF port to receive packets, and outputs a log message, for example, IRF member port GigabitEthernet1/4/0/1 does not work in receive direction, every 10 seconds. After that, if the IRF fabric splits, the IRF physical port can continue to work normally. If you disable IRF link failure detection or auto-recovery, the physical IRF port shuts down. To bring the port up, use the undo shutdown command. If IRF link auto-recovery is disabled, shuts down the physical IRF port. After that, the port stays in DOWN state, even if you enable IRF link auto-recovery function. (To bring the port up, use the undo shutdown command.) If the IRF fabric splits, the physical IRF ports in DOWN state change to the previous state.

Configuring MAD detection

You have the following MAD mechanisms for detecting multi-active collisions in different network scenarios: LACP MAD BFD MAD ARP MAD.

These MAD detection mechanisms operate independently, and you can configure all of them for an IRF fabric.

Configuring LACP MAD

1.

LACP MAD detection mechanism

23

With LACP MAD, an IRF member switch sends extended LACP data units (LACPDUs) with a type length value (TLV) that conveys the domain ID and active ID of the IRF fabric for detecting an IRF split. The domain ID uniquely identifies an IRF device in the network, and the active ID is identical to the member ID of the master switch in the IRF fabric. An IRF member switch compares the domain ID and the active ID in each received extended LACPDU with its domain ID and active ID: If the domain IDs are different, the extended LACPDU is from a different IRF fabric, and the switch does not continue to process the extended LACPDU with the MAD mechanism. If the domain IDs are the same, the switch compares the active IDs:

If the active IDs are different, the IRF fabric has split. If the active IDs are the same, the IRF fabric is operating normally.

2.

Network requirements

Every IRF member switch has a link with an intermediate switch, and all these links form a dynamic link aggregation group, as shown in Figure 9. The intermediate switch must be an H3C switch capable of handling extended LACPDUs that carry the Active ID field. For more information about LACP and the support of the switch for extended LACPDUs, see Layer 2LAN Switching Configuration Guide. CAUTION: If the intermediate switch is in an IRF fabric, you must assign this fabric a different domain ID than the LACP MAD-enabled fabric to avoid false detection of IRF partition.

24

Figure 9 Network diagram

3.

Configuring LACP MAD detection Create an aggregate interface (also required on the intermediate device); Configure the aggregation group to work in dynamic aggregation mode; (also required on the intermediate device) Enable LACP MAD detection on the dynamic aggregate interface; Add member ports to the aggregation group.(also required on the intermediate device)

Configure LACP MAD detection by following these steps:

To configure LACP MAD detection: To do

Enter system view Assign a domain ID to the IRF fabric Create an aggregate interface Configure the aggregation group to work in dynamic aggregation mode

Use the command

system-view irf domain domain-id interface bridge-aggregation interface-number

Remarks
Required By default, the domain ID of an IRF fabric is 0. Required Required

link-aggregation mode dynamic

By default, the aggregation group works in static aggregation mode.

25

To do

Use the command

Remarks
Required Disabled by default. This command can be configured on both static and dynamic aggregate interfaces, but it takes effect only on dynamic aggregate interfaces. This is because this detection approach depends on LACP. interface-type

Enable LACP MAD detection

mad enable

Return to system view Enter Ethernet interface view Assign the current Ethernet interface to the specified aggregation group

quit interface interface-number

port link-aggregation group number

Required

Configuring BFD MAD

1.

BFD MAD detection mechanism

BFD MAD is implemented with the BFD protocol. To configure BFD MAD detection, configure a MAD IP address on a BFD-enabled Layer 3 interface for each member switch. This MAD address identifies the member during BFD MAD detection. The MAD IP addresses assigned to the member switches must belong to the same network segment. When the IRF fabric operates normally, only the MAD IP address of the master is effective and the BFD session is down. You can use the display bfd session command to display the status of BFD sessions. The Session State field shows the session status. When the IRF fabric partitions, the MAD IP addresses of the masters in different IRF fabrics become effective to activate the BFD sessions to detect for multi-active IRF fabric collision. Network requirements

2.

If the IRF fabric has only two member switches, BFD MAD detection can work with or without intermediate switches. In a BFD MAD network without an intermediate switch, connect the IRF member switches with a dedicated BFD MAD link, as shown in Figure 10. The interfaces connected by the BFD MAD link must belong to the same VLAN. In VLAN interface view, assign different IP addresses on the same network segment for different member switches. If the IRF fabric has three or four member switches, there must be intermediate switches. In a BFD MAD network with an intermediate switch, connect each IRF member switch to the intermediate switch with a dedicated BFD MAD link, as shown in Figure 1 1. The interfaces connected by BFD MAD links must belong to the same VLAN, and you must assign different IP addresses on the same network segment for different member switches VLAN interfaces. CAUTION: A Layer 3 interface used for BFD MAD must be dedicated. Do not configure any other services on them. Otherwise, both the configured services and the BFD MAD detection function can be affected.

26

Figure 10 Network diagram (without intermediate switches)

27

Figure 11 Network diagram (with an intermediate switch)

3.

Configuring BFD MAD detection Create a VLAN dedicated to BFD MAD detection. This is also required on an intermediate switch, if it exists. Select the physical IRF ports for BFD MAD detection (at least one on each member switch) and add them into the detection-dedicated VLAN. This is also required on an intermediate switch, if it exists. Create VLAN interfaces for the detection-dedicated VLAN, enable BFD MAD detection on these interfaces, and then assign MAD IP addresses for them.

Configure BFD MAD detection in the following order:

To configure BFD MAD: To do

Enter system view Create a new VLAN dedicated to BFD MAD detection Return to system view Enter Ethernet interface view Assign the port Access port

Use the command

system-view vlan vlan-id quit interface interface-number port access vlan vlan-id interface-type

Remarks
Required The default VLAN on the switch is VLAN 1. Required

28

To do
to the VLAN dedicated to BFD MAD detection Trunk port

Use the command

port trunk permit vlan vlan-id

Remarks
You can select one approach according to the port type. BFD MAD detection has no requirement on the link type of the detection port, and you do not need to modify the current link type. By default, the port is an access port.

Hybrid port

port hybrid vlan vlan-id

Return to system view Enter VLAN interface view Enable BFD MAD Configure a MAD IP address for the VLAN interface on the specified member

quit interface interface-number mad bfd enable mad ip address ip-address { mask | mask-length } member member-id vlan-interface

Required Disabled by default. Required By default, no MAD IP address is configured for any interface.

NOTE: The VLAN enabled with BFD MAD detection must be dedicated to the BFD MAD detection and cannot be configured with any other service. A VLAN interface enabled with BFD MAD detection and the interfaces of this VLAN do not support any Layer 2 and Layer 3 protocol applications, including ARP and LACP. If the VLAN dedicated to BFD MAD detection contains a trunk port that permits the packets of multiple VLANs, make sure that the default VLAN of the trunk port is not the VLAN dedicated to BFD MAD detection. Otherwise, services configured on the trunk port may be affected. You cannot enable BFD MAD detection on VLAN-interface 1. The MAD function is mutually exclusive with the VPN. Layer 3 interfaces that have BFD MAD enabled cannot be bound to the VPN. Do not enable the spanning tree function on any port in a BFD MAD VLAN. The MAD function is mutually exclusive with the spanning tree function. You can assign the MAD IP address to an interface used for BFD MAD detection only with the mad ip address command. You cannot configure other IP addresses for it. This includes common IP address or VRRP virtual IP address configured with the ip address command.

Configuring ARP MAD

1.

ARP MAD detection mechanism

With ARP MAD, an IRF member switch sends extended gratuitous ARP packets that convey the domain ID and active ID of the IRF fabric for detecting an IRF split. The domain ID uniquely identifies an IRF fabric in the network, and the active ID is identical to the member ID of the master switch in the IRF fabric. An IRF member switch compares the domain ID and the active ID in each received extended gratuitous ARP packet with its domain ID and active ID: If the domain IDs are different, the extended gratuitous ARP packet is from a different IRF fabric, and the switch does not continue to process the packet with the MAD mechanism.
29

If the domain IDs are the same, the switch compares the active IDs:

If the active IDs are different, the IRF fabric has split. If the active IDs are the same, the IRF fabric is operating normally.

2.

Network requirements

Set up ARP MAD links between neighbor IRF member switches, or more commonly, between each IRF member switch and an intermediate switch (see Figure 12). If an intermediate switch is used, you must enable MSTP on the intermediate switch and the IRF fabric. CAUTION: If the intermediate device is in an IRF fabric, you must assign this fabric a different domain ID than the ARP MAD-enabled fabric to avoid false detection of IRF partition. Figure 12 Network diagram
Terminal network

Device

STP Domain To avoid loops, all devices in the domain have to be configured with the MSTP function.

IRF

IRF link Master Slave

Internet

Transmission path for common service packets Transmission path for gratuitous ARP packets

3.

Configuring ARP MAD detection Use the command

system-view irf domain domain-id

To configure ARP MAD: To do

Enter system view Assign a domain ID to the IRF fabric

Remarks
Required By default, the domain ID of an IRF fabric is 0.

30

To do
Create a new VLAN dedicated to ARP MAD detection Return to system view Enter Ethernet interface view Access port Trunk port Assign the port to the VLAN dedicated to ARP MAD detection

Use the command

vlan vlan-id quit interface interface-number port access vlan vlan-id port trunk permit vlan vlan-id interface-type

Remarks
Required The default VLAN on the switch is VLAN 1. Required You can select one approach according to the port type. ARP MAD detection has no requirement on the link type of the detection port, and you do not need to modify the current link type. By default, the port is an access port.

Hybrid port

port hybrid vlan vlan-id

Return to system view Enter VLAN interface view

quit interface interface-number vlan-interface

Required No IP address is assigned to any VLAN interface by default. Required By default, disabled. ARP MAD is

Assign the interface an IP address

ip address ip-address { mask | mask-length }

Enable ARP MAD

mad arp enable

Excluding a port from the shut down action upon detection of multi-active collision
By default, all service ports of an IRF fabric except the console and physical IRF ports shut down when the IRF fabric transits to recovery state upon detection of a multi-active collision. You can exclude a service port from the shut down action for management or other special purposes. For example, you can exclude a port from the shut down action, so you can telnet to the port for managing the switch. For another example, you can exclude a VLAN interface and its Layer 2 Ethernet interface from the shut down action, so you can log in through the VLAN interface. To configure a port not to shut down when the IRF fabric changes to recovery state: To do
Enter system view Configure a service port not to shut down when the IRF fabric transits to recovery state

Use the command

system-view mad exclude interface interface-type interface-number

Remarks
Required When an IRF fabric transits to recovery state, all its service ports are shut down by default.

31

Manually recovering an IRF fabric

An IRF link failure causes an IRF fabric to divide into two IRF fabrics and thus multi-active collision occurs. When the system detects the collision, it holds a role election between the two collided IRF fabrics. The IRF fabric whose masters member ID is smaller prevails and operates normally. The state of the other IRF fabric transits to recovery state and temporarily cannot forward data packets. In this case, recover the IRF fabric by repairing the IRF link first (the switch tries to automatically repair the failed IRF links. If the reparation fails, manually repair the failed links.) . When the link is recovered, the IRF fabric in active state and the IRF fabric in recovery state automatically merge into one. The IRF fabrics can merge into one only when you reboot the specified IRF fabric. If you reboot the IRF fabric in recovery state, two IRF fabrics merge into one. Service ports that were shut down and belonged to the IRF fabric in recovery state automatically restore their original physical state, and the whole IRF fabric recovers, as shown in Figure 13. If you reboot the IRF fabric in active state, two IRF fabrics merge into one. To enable the whole IRF fabric to recover in this case, execute the mad restore command to manually restore the original physical state of the service ports that were shut down and belonged to the IRF fabric in recovery state, as shown in Figure 14. Figure 13 Recover the IRF fabric when IRF link failure occurs
IP network IP network

IRF 1 (Active)

IRF 2 (Recovery)

IRF
After the IRF link is recovered

IP network

IP network

If the IRF fabric in active state fails due to exceptions (a member fails or link failure occurs, for example) before the IRF link is recovered, as shown in Figure 14, enable IRF fabric 2 (in recovery state) at the CLI by executing the mad restore command. Then, the state of IRF fabric 2 changes from recovery to active without the need of rebooting and takes over IRF fabric 1. Repair the IRF links. Then, the two IRF fabrics merge when the IRF link failure is recovered, and the IRF fabric recovers.

32

Figure 14 Recover the IRF fabric when the IRF link failure occurs and the IRF fabric in active state fails
IP network IP network

IRF 1 (Active)

IRF 2 (Recovery)
When the mad restore command is executed

IRF 2 (Active) IRF 1 fails due to physical problems

IP network

IP network

To manually recover an IRF fabric in recovery state: To do

Enter system view Recover an IRF fabric in recovery state

Use the command

system-view mad restore

Remarks
Required

Performing IRF configuration fast recovery

If a member switch has only one MPU and that board is corrupted, use the following fast recover function to avoid complicated IRF configuration.

Configuration prerequisites
H3C recommends that you perform the fast recovery operation when the switch operates in IRF mode. Save the IRF configuration file on all MPUs of the IRF fabric. Assume the name of the configuration file is a.cfg.

Configuration procedures

1. 2. 3.

Scenario I: Member switch 1 has one MPU, and it is corrupted. Member switch 2 has two MPUs. Modify the member ID of the standby MPU on member switch 2 to make it consistent with that of member switch 1. Unplug the corrupted MPU on member switch 1, and insert the standby MPU of member switch 2 into member switch 1. Configure IRF ports on member switch 1, and connect member switch 1 to member switch 2 through physical IRF ports. Scenario II: Both member switches have only one MPU, and the MPU on member switch 1 is corrupted.
33

The fast recovery is completed.

4. 5. 6. 7. 8.

Insert another MPU into member switch 2 and use it as the standby MPU of member switch 2. Copy the configuration file a.cfg on the active MPU of member switch 2 to the standby MPU, and set the file as the startup configuration file to be used at the next startup. Modify the member ID of the standby MPU on member switch 2 to make it consistent with that of member switch 1. Unplug the corrupted MPU on member switch 1, and insert the standby MPU of member switch 2 into member switch 1. Configure IRF ports on member switch 1, and connect member switch 1 to member switch 2 through physical IRF ports.

The fast recovery is completed. To modify the member ID of the standby MPU: To do Use the command
In IRF mode: Modify the member ID of the standby MPU in user view set irf chassis chassis-number slot-number member-id member-id In standalone mode: set irf slot member-id slot-number member-id slot

Remarks
This command is applicable to the fast recovery of the IRF only. If you use this command in other cases, unknown errors may occur. Use it with caution.

Displaying and maintaining an IRF fabric

To do
Display information about the IRF fabric Display topology about the IRF fabric information

Use the command

display irf [ | { begin | exclude | include } regular-expression ] display irf topology [ | { begin | exclude | include } regular-expression ] display irf configuration [ | { begin | exclude | include } regular-expression ] display mad [ verbose ] [ | { begin | exclude | include } regular-expression ] display restricted port [ chassis chassis-number slot slot-number ] [ | { begin | exclude | include } regular-expression ]

Remarks
Available in any view

Available in any view

Display all members configurations that take effect after device reboots Display configuration information about MAD detection

Available in any view

Available in any view

Available in any view A restricted port does not receive or forward multicast packets.

Display restricted ports

NOTE: A restricted port in the system does not receive or forward multicast packets.

IRF fabric configuration examples

34

NOTE: By default, Ethernet, VLAN, and aggregate interfaces are down. To configure these types of interfaces, execute the undo shutdown command to bring them up.

LACP MAD detection-enabled configuration example (non pre-configuration mode)

Network requirements
The network as shown in Figure 15 is outgrowing the forwarding capability of the existing core switch, (Device A). To address business growth, increase the number of ports at the access layer while protecting the current investments of the customer and maintaining ease of management and maintenance. Figure 15 Network diagram
IP network

GE4/0/1

GE4/0/2

Device C
GE1/4/0/2 XGE1/3/0/1 (IRF-port1/2) GE2/4/0/2 XGE 2/3/0/1 (IRF-port2/1)

Device A

IRF

Device B

Note: The solid orange line represents the IRF link; the solid magenta lines represent links used for LACP MAD; the solid black lines represent Ethernet links.

Configuration considerations
To increase the number of access ports, additional devices are needed. In this example, Device B is added. To offset the risk of IRF fabric partition, configure MAD to detect multi-active collisions. In this example, LACP MAD is adopted because there are many access switches. For LACP MAD, use an intermediate switch that supports extended LACPDUs.

Configuration procedure
1.

Configure Device A

# Change the operating mode of Device A to IRF.

<Sysname> system-view [Sysname] irf member 1 Info: Member ID change will take effect after the switch reboots and operates in IRF mode. [Sysname] chassis convert mode irf

35

The device will switch to IRF mode and reboot. You are recommended to save the current running configuration and specify the configuration file for the next startup. Continue? [Y/N]:y Do you want to convert the content of the next startup flash:/startup.cfg to make it available in IRF mode? [Y/N]:y Please wait... Saving the converted configuration file to the main board succeeded. Slot 1: Saving the converted configuration file succeeded. Now rebooting, please wait... configuration file

Device A reboots automatically to switch its operating mode. # Create IRF port 1/2 and bind it with the physical IRF port Ten-GigabitEthernet 1/3/0/1.
<Sysname> system-view [Sysname] irf-port 1/2 [Sysname-irf-port 1/2] port group interface Ten-Gigabitethernet 1/3/0/1 [Sysname-irf-port 1/2] save The current configuration will be written to the device. Are you sure? [Y/N]:y Please input the file name(*.cfg)[flash:/cfa0555.cfg] (To leave the existing filename unchanged, press the enter key): flash:/cfa0555.cfg exists, overwrite? [Y/N]:y Validating file. Please wait..................................... The current configuration is saved to the active main board successfully. Configuration is saved to device successfully.

2.

Configure Device B

# Change the operating mode of Device B to IRF.

<Sysname> system-view [Sysname] irf member 2 Info: Member ID change will take effect after the switch reboots and operates in IRF mode. [Sysname] chassis convert mode irf The device will switch to IRF mode and reboot. You are recommended to save the current running configuration and specify the configuration file for the next startup. Continue? [Y/N]:y Do you want to convert the content of the next startup flash:/startup.cfg to make it available in IRF mode? [Y/N]:y Please wait... Saving the converted configuration file to the main board succeeded. Slot 1: Saving the converted configuration file succeeded. Now rebooting, please wait... configuration file

Device B reboots automatically to switch its operating mode. # Create IRF port 2/1 and bind it with the physical IRF port Ten-GigabitEthernet 2/3/0/1.
<Sysname> system-view [Sysname] irf-port 2/1 [Sysname-irf-port 2/1] port group interface Ten-Gigabitethernet 2/3/0/1 [Sysname-irf-port 2/1] save The current configuration will be written to the device. Are you sure? [Y/N]:y Please input the file name(*.cfg)[flash:/cfa0666.cfg]

36

(To leave the existing filename unchanged, press the enter key): flash:/cfa0666.cfg exists, overwrite? [Y/N]:y Validating file. Please wait..................................... The current configuration is saved to the active main board successfully. Configuration is saved to device successfully.

3. 4.

Connect the two devices as shown in Figure 15 with IRF cables, and then reboot Device B. Power them on, and the IRF is established. Configure LACP MAD

# Set the domain ID of the IRF fabric to 1.

<Sysname> system-view [Sysname] irf domain 1

# Create a dynamic aggregate interface and enable LACP MAD.

<Sysname> system-view [Sysname] interface bridge-aggregation 2 [Sysname-Bridge-Aggregation2] link-aggregation mode dynamic [Sysname-Bridge-Aggregation2] mad enable You need to assign a domain ID (range: 0-4294967295) [Current domain is: 1]: The assigned domain ID is: 1 Info: MAD LACP only enable on dynamic aggregation interface [Sysname-Bridge-Aggregation2] quit

# Assign ports 1/4/0/2 and 2/4/0/2 to the aggregate interface for LACP MAD detection.
Sysname] interface Gigabitethernet 1/4/0/2 [Sysname-GigabitEthernet1/4/0/2] port link-aggregation group 2 [Sysname-GigabitEthernet1/4/0/2] quit [Sysname] interface Gigabitethernet 2/4/0/2 [Sysname-GigabitEthernet2/4/0/2] port link-aggregation group 2

5.

Configure Device C as the intermediate device

Acting as the intermediate device, Device C forwards and processes extended LACPDUs sent between Device A and Device B for MAD detection. CAUTION: If the intermediate device is in an IRF fabric, you must assign this fabric a different domain ID than the LACP MAD-enabled fabric. # Create a dynamic aggregate interface.
<Sysname> system-view [Sysname] interface bridge-aggregation 2 [Sysname-Bridge-Aggregation2] link-aggregation mode dynamic [Sysname-Bridge-Aggregation2] quit

# Assign ports GigabitEthernet 4/0/1 and GigabitEthernet 4/0/2 to the dynamic aggregate interface.
[Sysname] interface Gigabitethernet 4/0/1 [Sysname-GigabitEthernet4/0/1] port link-aggregation group 2 [Sysname-GigabitEthernet4/0/1] quit [Sysname] interface Gigabitethernet 4/0/2 [Sysname-GigabitEthernet4/0/2] port link-aggregation group 2

37

NOTE: On the switch that supports pre-configurations, you can also configure the LACP MAD; on the switch not supporting pre-configurations, you can also configure the BFD MAD.

BFD MAD detection-enabled IRF configuration example (pre-configuration mode)

Network requirements
The network as shown in Figure 16 is outgrowing the forwarding capability of the existing core switch (Device A). To address business growth, scale up the network to extend its forwarding capability while protecting the present investments of the customer. Ease of management and maintenance must also be ensured. Figure 16 Network diagram

Configuration considerations
Device A is located at the distribution layer of the network. To improve the forwarding capability at this layer, additional devices are needed. In this example, Device B is added. To offset the risk of IRF fabric partition, configure MAD to detect multi-active collisions. In this example, BFD MAD is adopted.

Configuration procedure
1.

Configure Device A

# Set the member ID of Device A to 1, priority to 12, create IRF port 2, and bind it to physical port Ten-GigabitEthernet 3/0/1.
38

<Sysname> system-view [Sysname] irf member 1 Info: Member ID change will take effect after the switch reboots and operates in IRF mode. [Sysname] irf priority 12 [Sysname] irf-port 2 [Sysname-irf-port 2] port group interface ten-gigabitethernet 3/0/1 [Sysname-irf-port 2] quit

# Save the running configuration to the configuration file for next startup.
[Sysname] save The current configuration will be written to the device. Are you sure? [Y/N]:y Please input the file name(*.cfg)[flash:/cfa0516.cfg] (To leave the existing filename unchanged, press the enter key): flash:/cfa0516.cfg exists, overwrite? [Y/N]:y Validating file. Please wait..................................... The current configuration is saved to the active main board successfully. Configuration is saved to device successfully.

# Change the operating mode of Device A to IRF.

<Sysname> system-view [Sysname] chassis convert mode irf The device will switch to IRF mode and reboot. You are recommended to save the current running configuration and specify the configuration file for the next startup. Continue? [Y/N]:y Do you want to convert the content of the next startup flash:/startup.cfg to make it available in IRF mode? [Y/N]:y Please wait... Saving the converted configuration file to the main board succeeded. Slot 1: Saving the converted configuration file succeeded. Now rebooting, please wait... configuration file

Device A operates in IRF mode after reboot.

2.

Configure Device B

# Set the member ID of Device B to 2, priority to 26, create IRF port 1, and bind it to physical port Ten-GigabitEthernet 3/0/1.
<Sysname> system-view [Sysname] irf member 2 Info: Member ID change will take effect after the switch reboots and operates in IRF mode. [Sysname] irf priority 26 [Sysname] irf-port 1 [Sysname-irf-port 1] port group interface ten-gigabitethernet 3/0/1 [Sysname-irf-port 1] quit

# Save the running configuration to the configuration file for next startup.
[Sysname] save The current configuration will be written to the device. Are you sure? [Y/N]:y Please input the file name(*.cfg)[flash:/cfa0516.cfg] (To leave the existing filename unchanged, press the enter key): flash:/cfa0516.cfg exists, overwrite? [Y/N]:y Validating file. Please wait.....................................

39

The current configuration is saved to the active main board successfully. Configuration is saved to device successfully.

# Connect the two devices as shown in Figure 16. # Switch the operating mode of Device B to IRF.
<Sysname> system-view [Sysname] chassis convert mode irf The device will switch to IRF mode and reboot. You are recommended to save the current running configuration and specify the configuration file for the next startup. Continue? [Y/N]:y

Do you want to convert the content of the next startup flash:/startup.cfg to make it available in IRF mode? [Y/N]:y Please wait...

configuration

file

Saving the converted configuration file to the main board succeeded. Slot 1: Saving the converted configuration file succeeded. Now rebooting, please wait...

Device B and Device A form an IRF fabric when Device B reboots.

3.

Configure BFD MAD

# Create VLAN 3, and add port GigabitEthernet 1/4/0/1 (located on Device A) and port GigabitEthernet 2/4/0/1 (located on Device B) to VLAN 3.
<Sysname> system-view [Sysname] vlan 3 [Sysname-vlan3] port gigabitethernet 1/4/0/1 gigabitethernet 2/4/0/1 [Sysname-vlan3] quit

# Create VLAN-interface 3 and configure a MAD IP address for the interface.

[Sysname] interface vlan-interface 3 [Sysname-Vlan-interface3] mad bfd enable [Sysname-Vlan-interface3] mad ip address 192.168.2.1 24 member 1 [Sysname-Vlan-interface3] mad ip address 192.168.2.2 24 member 2 [Sysname-Vlan-interface3] quit

# Because BFD MAD detection and spanning tee function are mutually exclusive, disable the spanning tree function on ports GigabitEthernet 1/4/0/1 and GigabitEthernet 2/4/0/1.
[Sysname] interface Gigabitethernet 1/4/0/1 [Sysname-Gigabitethernet1/4/0/1] undo stp enable [Sysname-Gigabitethernet1/4/0/1] quit [Sysname] interface Gigabitethernet 2/4/0/1 [Sysname-Gigabitethernet2/4/0/1] undo stp enable

ARP MAD detection-enabled IRF configuration example (pre-configuration mode)

Network requirements
The network (see Figure 17) is outgrowing the forwarding capability of the existing core switch, namely, Device A.
40

To address business growth, scale up the network to extend its forwarding capability while protecting the present investments of the customer. Ease of management and maintenance must also be ensured. Figure 17 Network diagram

Configuration considerations
Device A is located at the distribution layer of the network. To improve the forwarding capability at this layer, additional devices are needed. In this example, Device B is added. To offset the risk of IRF fabric partition, configure MAD to detect multi-active collisions. In this example, ARP MAD is adopted, and the ARP MAD packets are transmitted over dual links connected to Device C. Enable the spanning tree function on the IRF fabric and Device to prevent loops.

Configuration procedure
1.

Configure Device A

# Set the member ID of Device A to 1, create IRF port 2, and bind it to physical port Ten-GigabitEthernet 3/0/1.
<Sysname> system-view [Sysname] irf member 1 Info: Member ID change will take effect after the switch reboots and operates in IRF mode. [Sysname] irf-port 2 [Sysname-irf-port2] port group interface ten-gigabitethernet 3/0/1 [Sysname-irf-port2] quit

# Save the running configuration to the configuration file for next startup.
[Sysname] save The current configuration will be written to the device. Are you sure? [Y/N]:y Please input the file name(*.cfg)[flash:/cfa0516.cfg] (To leave the existing filename unchanged, press the enter key): flash:/cfa0516.cfg exists, overwrite? [Y/N]:y Validating file. Please wait..................................... The current configuration is saved to the active main board successfully. Configuration is saved to device successfully.

# Change the operating mode of Device A to IRF.

41

<Sysname> system-view [Sysname] chassis convert mode irf The device will switch to IRF mode and reboot. You are recommended to save the current running configuration and specify the configuration file for the next startup. Continue? [Y/N]:y Do you want to convert the content of the next startup flash:/startup.cfg to make it available in IRF mode? [Y/N]:y Please wait... Saving the converted configuration file to the main board succeeded. Slot 1: Saving the converted configuration file succeeded. Now rebooting, please wait... configuration file

Device A operates in IRF mode after reboot.

2.

Configure Device B

# Set the member ID of Device B to 2, create IRF port 1, and bind it to physical port Ten-GigabitEthernet 3/0/1.
<Sysname> system-view [Sysname] irf member 2 Info: Member ID change will take effect after the switch reboots and operates in IRF mode. [Sysname] irf-port 1 [Sysname-irf-port1] port group interface ten-gigabitethernet 3/0/1 [Sysname-irf-port1] quit

# Save the running configuration to the configuration file for next startup.
[Sysname] save The current configuration will be written to the device. Are you sure? [Y/N]:y Please input the file name(*.cfg)[flash:/cfa0516.cfg] (To leave the existing filename unchanged, press the enter key): flash:/cfa0516.cfg exists, overwrite? [Y/N]:y Validating file. Please wait..................................... The current configuration is saved to the active main board successfully. Configuration is saved to device successfully.

# Connect the two devices as shown in Figure 17. # Switch the operating mode of Device B to IRF.
<Sysname> system-view [Sysname] chassis convert mode irf The device will switch to IRF mode and reboot. You are recommended to save the current running configuration and specify the configuration file for the next startup. Continue? [Y/N]:y Do you want to convert the content of the next startup flash:/startup.cfg to make it available in IRF mode? [Y/N]:y Please wait... Saving the converted configuration file to the main board succeeded. Slot 1: Saving the converted configuration file succeeded. Now rebooting, please wait... configuration file

Device B and Device A form an IRF after Device B reboots.

3.

Configure the IRF fabric

42

# Set the domain ID of the IRF fabric to 1.

[Sysname] irf domain 1

# Enable the spanning tree function globally on the IRF fabric to prevent loops.
<Sysname> system-view [Sysname] stp enable

# Configure that the bridge MAC address of the IRF fabric changes as soon as the master leaves.
[Sysname] undo irf mac-address persistent

4.

Configure ARP MAD

# Create VLAN 3, and add port GigabitEthernet 1/4/0/2 (located on Device A) and port GigabitEthernet 2/4/0/2(located on Device B) to VLAN 3.
<Sysname> system-view [Sysname] vlan 3 [Sysname-vlan3] port gigabitethernet 1/4/0/2 gigabitethernet 2/4/0/2 [Sysname-vlan3] quit

# Create VLAN-interface 3, assign it an IP address, and enable ARP MAD on the interface.
[Sysname] interface vlan-interface 3 [Sysname-Vlan-interface3] mad arp enable You need to assign a domain ID (range: 0-4294967295) [Current domain is: 1]: The assigned domain ID is: 1 [Sysname-Vlan-interface3] ip address 192.168.2.1 24

5.

Configure Device C as the intermediate device

Acting as the intermediate device, Device C forwards and processes gratuitous ARP packets sent between Device A and Device B for MAD detection. CAUTION: If the intermediate device is in an IRF fabric, you must assign this fabric a different domain ID than the ARP MAD-enabled fabric. # Enable the spanning tree function globally on Device C to prevent loops.
<DeviceC> system-view [DeviceC] stp enable

# Create VLAN 3, and add port GigabitEthernet 4/0/1 and port GigabitEthernet 4/0/2 to VLAN 3 to forward ARP MAD packets.
[DeviceC] vlan 3 [DeviceC-vlan3] port gigabitethernet 4/0/1 gigabitethernet 4/0/2 [DeviceC-vlan3] quit

Switching the operating mode of IRF member switches from IRF to standalone
Network requirements
As shown in Figure 18, the IRF fabric operates normally. Device A and Device B are IRF member switches. To adapt the network to changes, switch the operating mode of Device A and Device B from IRF to standalone.
43

Figure 18 Network diagram

Configuration considerations
Disconnect IRF connections. Unplug cables for the IRF connections or shut down all physical IRF ports on the master at the CLI. This example uses the latter method. After the IRF is partitioned, switch the operating mode of the two switches from IRF to standalone respectively.

Configuration procedure
1.

Find out the master.

Slot 0 1 0 1 Role Master Slave Slave Slave Priority 1 1 1 1 CPU-Mac 00e0-fc0a-15e0 00e0-fc0f-8c02 00e0-fc0f-15e1 00e0-fc0f-15e2 Description DeviceA DeviceA DeviceB DeviceB

<IRF> display irf Switch *+1 1 2 2

--------------------------------------------------

* indicates the device is the master. + indicates the device through which the user logs in.

The Bridge MAC of the IRF is: 000f-e26a-58ed Auto upgrade Mac persistent Domain ID : no : always : 0

The output shows that Device A is the master.

2.

Disconnect IRF connections: disable the physical IRF port Ten-Gigabitethernet 1/3/0/1 on the master at the CLI. (In this example, there is only one IRF physical link.)

<IRF> system-view [IRF] interface ten-gigabitethernet 1/3/0/1 [IRF-Ten-Gigabitethernet1/3/0/1] shutdown [IRF-Ten-Gigabitethernet1/3/0/1] quit

44

3.

Switch the operating mode of Device A from IRF to standalone.

[IRF] undo chassis convert mode The device will switch to stand-alone mode and reboot. You are recommended to save the current running configuration and specify the configuration file for the next startup. Continue? [Y/N]:y Do you want to convert the content of the next startup configuration file flash:/vrpcfg.cfg to make it available in stand-alone mode? [Y/N]:y Please wait............. Saving the converted configuration file to main board succeeded. Chassis 1 Slot 1: Saving the converted configuration file succeeded. Now rebooting, please wait...

Device A automatically reboots to complete the operating mode switch.

4.

Log in to Device B and switch the operating mode of Device B from IRF to standalone.

<IRF> system-view [IRF] undo chassis convert mode The device will switch to stand-alone mode and reboot. You are recommended to save the current running configuration and specify the configuration file for the next startup. Continue? [Y/N]:y Do you want to convert the content of the next startup configuration file flash:/vrpcfg.cfg to make it available in stand-alone mode? [Y/N]:y Please wait............. Saving the converted configuration file to main board succeeded. Chassis 2 Slot 1: Saving the converted configuration file succeeded. Now rebooting, please wait...

Device B automatically reboots to complete the operating mode switch. NOTE: If you have created VLAN interfaces and configured IP addresses for the interfaces on the IRF fabric, and both Device A and Device B have member ports of this VLAN, IP address collision occurs after the operating mode of Device A and Device B switches to standalone. To remove the collision, log in to one device and modify the IP address of this VLAN.

Enhanced IRF mode configuration example (four switches forming an IRF fabric)
Network requirements
As shown in Figure 19, Device A cannot provide enough access bandwidth. To increase access bandwidth, add three devices and configure them and Device A to form an IRF fabric, as shown in Figure 20.

45

Figure 19 Network diagram before IRF configuration

IP network
Device E

Device A

Figure 20 Network diagram for IRF configuration

Device E

IP network

XGE1/3/0/2 (IRF-port1/1)

XGE1/3/0/1 (IRF-port1/2)

XGE2/3/0/1 (IRF-port2/1)

XGE3/3/0/2 (IRF-port3/1)

XGE3/3/0/1 (IRF-port3/2) XGE4/3/0/1 (IRF-port4/1)

XGE4/3/0/2 (IRF-port4/2)

Device A

Device B

XGE2/3/0/2 (IRF-port2/2)

Device C

Device D

The orange lines represent IRF links. The black lines represent Ethernet links.

Configuration considerations
Configure the member IDs, priorities, and IRF ports on the four member switches.
46

 Enable enhanced IRF mode on the four member switches, and connect them as shown in Figure 20. Change the operating mode to IRF on the four member switches.

Configuration procedure
1.

Configure Device A.

# Set the member ID to 1, and priority to 12.

<Sysname> system-view [Sysname] irf member 1 Info: Member ID change will take effect after the switch reboots and operates in IRF mode. [Sysname] irf priority 12

# Create IRF port 1, and bind it to physical port Ten-GigabitEthernet 3/0/2.

[Sysname] irf-port 1 [Sysname-irf-port 1] port group interface ten-gigabitethernet 3/0/2 [Sysname-irf-port 1] quit

# Create IRF port 2, and bind it to physical port Ten-GigabitEthernet 3/0/1.

[Sysname] irf-port 2 [Sysname-irf-port 2] port group interface ten-gigabitethernet 3/0/1 [Sysname-irf-port 2] quit

# Enable enhanced IRF mode.

[Sysname] irf mode enhanced

# Save the configuration to the configuration file to be used at the next startup.
[Sysname] save The current configuration will be written to the device. Are you sure? [Y/N]:y Please input the file name(*.cfg)[flash:/cfa0511.cfg] (To leave the existing filename unchanged, press the enter key): flash:/cfa0511.cfg exists, overwrite? [Y/N]:y Validating file. Please wait..................................... The current configuration is saved to the active main board successfully. Configuration is saved to device successfully

# Change the operating mode to IRF.

<Sysname> system-view [Sysname] chassis convert mode irf The device will switch to IRF mode and reboot. You are recommended to save the current running configuration and specify the configuration file for the next startup. Continue? [Y/N]:y Do you want to convert the content of the next startup flash:/startup.cfg to make it available in IRF mode? [Y/N]:y Please wait... Saving the converted configuration file to the main board succeeded. Slot 1: Saving the converted configuration file succeeded. Now rebooting, please wait... configuration file

After the reboot, an IRF fabric comprising only Device A is created.

2.

Configure Device B

# Set the member ID to 2, and priority to 26.

47

<Sysname> system-view [Sysname] irf member 2 Info: Member ID change will take effect after the switch reboots and operates in IRF mode. [Sysname] irf priority 26

# Create IRF port 1, and bind it to physical port Ten-GigabitEthernet 3/0/1.

[Sysname] irf-port 1 [Sysname-irf-port 1] port group interface ten-gigabitethernet 3/0/1 [Sysname-irf-port 1] quit

# Create IRF port 2, and bind it to physical port Ten-GigabitEthernet 3/0/2.

[Sysname] irf-port 2 [Sysname-irf-port 2] port group interface ten-gigabitethernet 3/0/2 [Sysname-irf-port 2] quit

# Enable enhanced IRF mode.

[Sysname] irf mode enhanced

# Save the configuration to the configuration file to be used at the next startup.
[Sysname] save The current configuration will be written to the device. Are you sure? [Y/N]:y Please input the file name(*.cfg)[flash:/cfa0512.cfg] (To leave the existing filename unchanged, press the enter key): flash:/cfa0512.cfg exists, overwrite? [Y/N]:y Validating file. Please wait..................................... The current configuration is saved to the active main board successfully. Configuration is saved to device successfully

# Connect the switches as shown in Figure 20. # Change the operating mode to IRF.
<Sysname> system-view [Sysname] chassis convert mode irf The device will switch to IRF mode and reboot. You are recommended to save the current running configuration and specify the configuration file for the next startup. Continue? [Y/N]:y Do you want to convert the content of the next startup flash:/startup.cfg to make it available in IRF mode? [Y/N]:y Please wait... Saving the converted configuration file to the main board succeeded. Slot 1: Saving the converted configuration file succeeded. Now rebooting, please wait... configuration file

After the reboot, the IRF fabric comprises Device A and Device B.
3.

Configure Device C.

# Set the member ID to 3, and priority to 6.

<Sysname> system-view [Sysname] irf member 3 Info: Member ID change will take effect after the switch reboots and operates in IRF mode. [Sysname] irf priority 6

# Create IRF port 1, and bind it to physical port Ten-GigabitEthernet 3/0/2.

48

[Sysname] irf-port 1 [Sysname-irf-port 1] port group interface ten-gigabitethernet 3/0/2 [Sysname-irf-port 1] quit

# Create IRF port 2, and bind it to physical port Ten-GigabitEthernet 3/0/1.

[Sysname] irf-port 2 [Sysname-irf-port 2] port group interface ten-gigabitethernet 3/0/1 [Sysname-irf-port 2] quit

# Enable enhanced IRF mode.

[Sysname] irf mode enhanced

# Save the configuration to the configuration file to be used at the next startup.
[Sysname] save The current configuration will be written to the device. Are you sure? [Y/N]:y Please input the file name(*.cfg)[flash:/cfa0513.cfg] (To leave the existing filename unchanged, press the enter key): flash:/cfa0513.cfg exists, overwrite? [Y/N]:y Validating file. Please wait..................................... The current configuration is saved to the active main board successfully. Configuration is saved to device successfully

# Connect the switches as shown in Figure 20. # Change the operating mode to IRF.
<Sysname> system-view [Sysname] chassis convert mode irf The device will switch to IRF mode and reboot. You are recommended to save the current running configuration and specify the configuration file for the next startup. Continue? [Y/N]:y Do you want to convert the content of the next startup flash:/startup.cfg to make it available in IRF mode? [Y/N]:y Please wait... Saving the converted configuration file to the main board succeeded. Slot 1: Saving the converted configuration file succeeded. Now rebooting, please wait... configuration file

After the reboot, the IRF fabric comprises Device A, Device B, and Device C.
4.

Configure Device D.

# Set the member ID to 4, and priority to 2.

<Sysname> system-view [Sysname] irf member 4 Info: Member ID change will take effect after the switch reboots and operates in IRF mode. [Sysname] irf priority 2

# Create IRF port 1, and bind it to physical port Ten-GigabitEthernet 3/0/1.

[Sysname] irf-port 1 [Sysname-irf-port 1] port group interface ten-gigabitethernet 3/0/1 [Sysname-irf-port 1] quit

# Create IRF port 2, and bind it to physical port Ten-GigabitEthernet 3/0/2.

[Sysname] irf-port 2

49

[Sysname-irf-port 2] port group interface ten-gigabitethernet 3/0/2 [Sysname-irf-port 2] quit

# Enable enhanced IRF mode.

[Sysname] irf mode enhanced

# Save the configuration to the configuration file to be used at the next startup.
[Sysname] save The current configuration will be written to the device. Are you sure? [Y/N]:y Please input the file name(*.cfg)[flash:/cfa0514.cfg] (To leave the existing filename unchanged, press the enter key): flash:/cfa0514.cfg exists, overwrite? [Y/N]:y Validating file. Please wait..................................... The current configuration is saved to the active main board successfully. Configuration is saved to device successfully

# Connect the switches as shown in Figure 20. # Change the operating mode to IRF.
<Sysname> system-view [Sysname] chassis convert mode irf The device will switch to IRF mode and reboot. You are recommended to save the current running configuration and specify the configuration file for the next startup. Continue? [Y/N]:y Do you want to convert the content of the next startup flash:/startup.cfg to make it available in IRF mode? [Y/N]:y Please wait... Saving the converted configuration file to the main board succeeded. Slot 1: Saving the converted configuration file succeeded. Now rebooting, please wait... configuration file

After the reboot, the IRF fabric comprises Device A, Device B, Device C, and Device D.

50

Index
ABCDEIPS
A Accessing an IRF fabric,15 B Basic concepts,2 C Configuring IRF member switches in IRF mode,16 D Displaying and maintaining an IRF fabric,34 E Enabling enhanced IRF mode,13 Establishment, operation, and maintenance of an IRF fabric,4 I IRF fabric configuration examples,34 IRF fabric configuration task list,8 IRF overview,1 P Performing IRF configuration fast recovery,33 Pre-configuring an IRF member switch in standalone mode,1 1 S Saving the running configuration to the configuration file to be used at the next system startup,14 Switching operating mode,14