H3C S6820 Series Quick guide

H3C S6820 Switches
IRF Configuration Examples
Copyright © 2020 New H3C Technologies Co., Ltd. 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 New
H3C Technologies Co., Ltd.
Except for the trademarks of New H3C Technologies Co., Ltd., any trademarks that may be mentioned in this document are
the property of their respective owners.
The information in this document is subject to change without notice.

i
Contents
Introduction·····················································································1
Prerequisites···················································································1
General restrictions and guidelines ······················································1
Hardware requirements···············································································································1
Software requirements················································································································1
Feature compatibility and configuration restrictions············································································1
IRF physical interface requirements·······························································································2
Connecting IRF ports··················································································································3
Licensing requirements ···············································································································3
Example: Setting up a four-chassis LACP MAD-enabled IRF fabric·············3
Network configuration ·················································································································3
Analysis···································································································································4
IRF setup···························································································································4
LACP MAD configuration ······································································································6
Network connectivity configuration ··························································································6
Software versions used···············································································································7
Restrictions and guidelines ··········································································································7
Restrictions and guidelines for LACP MAD configuration ·····························································7
Restrictions and guidelines for IRF physical interfaces ································································7
Procedures·······························································································································7
Setting up the access-layer IRF fabric······················································································7
Configuring LACP MAD······································································································10
Configuring network connectivity settings ···············································································12
Verifying the configuration··········································································································13
Verifying the IRF setup·······································································································13
Verifying the link backup function of multichassis aggregations ···················································13
Verifying link failure protection of the ring topology ···································································14
Verifying the LACP MAD configuration···················································································14
Configuration files ····················································································································16
Example: Setting up a four-chassis BFD MAD-enabled IRF fabric ············20
Network configuration ···············································································································20
Analysis·································································································································22
IRF setup·························································································································22
BFD MAD configuration······································································································23
Network connectivity configuration ························································································24
Software versions used·············································································································25
Restrictions and guidelines ········································································································26
Restrictions and guidelines for BFD MAD configuration ·····························································26
IRF port binding ················································································································26
Procedures·····························································································································26
Setting up the IRF fabric ·····································································································26
Configuring BFD MAD········································································································29
Configuring network connectivity settings ···············································································30
Verifying the configuration··········································································································35
Verifying the IRF setup·······································································································35
Verifying the routing configuration·························································································35
Verifying the link backup function of multichassis aggregations ···················································36
Verifying link failure protection of the ring topology ···································································36
Verifying the BFD MAD configuration·····················································································36
Configuration files ····················································································································38
Related documentation····································································46

1
Introduction
This document provides examples for setting up IRF fabrics and configuring link aggregation and
routing on the IRF fabrics.
Prerequisites
The configuration examples in this document were created and verified in a lab environment, and all
the devices were started with the factory default configuration. When you are working on a live
network, make sure you understand the potential impact of every command on your network.
This document assumes that you have basic knowledge of IRF.
General restrictions and guidelines
When you set up and configure an IRF fabric, follow the restrictions and guidelines in this section.
This section provides only the basic restrictions and guidelines that ensure a successful IRF
deployment. For complete information, see IRF configuration guide for the switch.
Hardware requirements
An S6820 switch can form an IRF fabric only with switches in the same series.
Software requirements
All IRF member devices must run the same system software version.
Feature compatibility and configuration
restrictions
Make sure the feature settings in Table 1 are the same across member devices.
Table 1 IRF and feature compatibility
Feature Command
Enhanced ECMP mode. ecmp mode enhanced
Support for the IPv6 routes with prefixes longer
than 64 bits. hardware-resource routing-mode ipv6-128
Table capacity mode. hardware-resource switch-mode
VXLAN hardware resource mode. hardware-resource vxlan
Maximum number of ECMP routes max-ecmp-num
Enabling status of the packet loss prevention
feature for OpenFlow forwarding. openflow lossless enable
System operating mode. system-working-mode

2
IRF physical interface requirements
Candidate IRF physical interfaces
Use the following ports as IRF physical interfaces:
•SFP28 ports operating at 25 Gbps.
•QSFP+ ports operating at 40 Gbps.
•QSFP28 ports operating at 40 Gbps or 100 Gbps, or their 25-GE breakout interfaces.
The ports on LSWM18CQMSEC modules cannot be used as IRF physical interfaces.
Grouped port binding restrictions
The S6820 switch series contains grouped ports, as shown in Table 2.
When you use the ports in a group for IRF links, follow these restrictions and guidelines:
•If you use one port in a group as an IRF physical interface, the remaining ports in the group can
only act as IRF physical interfaces. You cannot use them for any other purposes. If you use only
some of them for IRF connection, the remaining ports can only be idle.
•The ports in a group can be bound to different IRF ports.
•Before you bind one port to an IRF port, you must shut down all the ports that have not been
bound to IRF ports. If any of the unbound ports are in up state, the bind action will fail.
•Before you remove one port from an IRF port, you must shut down all the ports. If any of the
ports are in up state, the remove action will fail.
•Bring up the ports after you complete the bind operation. If a port is not bound to an IRF port, the
system prevents you from bringing up the port so it cannot be used for purposes other than IRF
connection.
NOTE:
These restrictions on gouped ports do not apply if you use 100-GE ports that are not split or use
40-GE ports (except ports on the LSWM116Q module) for IRF connections.
Table 2 Port grouping rules
Hardware Port grouping rules
The S6820-56HF switch
The switch has two fixed port ranges. Ports numbered
1 to 24 are in one range and ports numbered 33 to 56
are in the other range. In each range, the ports are
divided into four-port groups by port number in order,
starting from the lowest number.
The LSWM124TG2H module The ports numbered 1 to 24 are divided into four-port
groups by port number in order, starting from 1.
The LSWM116Q module
Odd and even numbered ports are grouped separately
into two-port groups by port number, starting from the
lowest odd or even number. Each group contains two
ports with consecutive odd or even port numbers. For
example, ports 1 and 3 are in one group, and ports 2
and 4 are in another group.
All S6820 switches The four 25-GE breakout interfaces of a 100-GE port
are in one port group. To split a 100-GE port, use the
using twenty-fivegige command.

3
Connecting IRF ports
When you connect two neighboring IRF members, connect the physical interfaces of IRF-port 1 on
one member to the physical interfaces of IRF-port 2 on the other.
Licensing requirements
For a license-based feature to run correctly on an IRF fabric, make sure the licenses installed for the
feature on all member devices are the same. For more information about feature licensing, see
Fundamentals Configuration Guide.
Example: Setting up a four-chassis LACP
MAD-enabled IRF fabric
Network configuration
As shown in Figure 2, use a four-chassis IRF fabric to replace the ToR switches in Figure 1 at the
access layer of the data center. The access-layer IRF fabric provides Layer 2 forwarding services.
Run LACP MAD to detect IRF split.
Figure 1 Network diagram before IRF deployment
Distribution
Access
Rack Rack

4
Figure 2 Network diagram after IRF deployment
Analysis
The requirements in this example include the following categories:
•IRF setup
•LACP MAD configuration
•Network connectivity configuration
IRF setup
To set up an IRF fabric, determine the items in Table 3.
Table 3 Basic IRF setup
Item Analysis Choice in this example
Topology You can use a ring or daisy-chain topology for a
three- or four-chassis IRF fabric.
For reliability, use the ring topology. Ring topology (see Figure 3).
Member ID
assignment IRF member IDs must be unique.
Device A—1.
Device B—2.
Device C—3.
Device D—4.
Master device IRF members elect a master automatically. For a member device to be elected the master, assign it
the highest member priority. Device A.
IRF port bindings
Make sure the IRFport bindings are consistent with
the physical connections.
For high availability, bind multiple physical
interfaces to an IRF port. These ports will
automatically aggregate for load balancing and
redundancy.
See Table 4.
IRF
Distribution
Access

5
Figure 3 IRF fabric topology
Table 4 IRF physical interface bindings
IRF port IRF physical interfaces
Device A:
IRF-port 1 HundredGigE 1/0/25
HundredGigE 1/0/26
IRF-port 2 HundredGigE 1/0/27
HundredGigE 1/0/28
Device B:
IRF-port 1 HundredGigE 2/0/25
HundredGigE 2/0/26
IRF-port 2 HundredGigE 2/0/27
HundredGigE 2/0/28
Device C:
IRF-port 1 HundredGigE 3/0/25
HundredGigE 3/0/26
IRF-port 2 HundredGigE 3/0/27
HundredGigE 3/0/28
Device D:
IRF-port 1 HundredGigE 4/0/25
HundredGigE 4/0/26
IRF-port 2 HundredGigE 4/0/27
HundredGigE 4/0/28
NOTE:
The first segment in a physical interface number is the IRF member ID. By default, the IRF membe
r
ID is 1. Table 4 shows the interface numbers after the member IDs are changed.

6
LACP MAD configuration
IMPORTANT:
For LACP MAD to run correctly, you must make sure the intermediate device supports extended
LACPDUs for LACP MAD.
To run LACP MAD, the IRF fabric and intermediate device must each have a dynamic Ethernet link
aggregation. LACP MAD cannot run on a static link aggregation.
If the intermediate device is also an IRF fabric, you must configure MAD for both IRF fabrics, and
assign the two IRF fabrics different domain IDs for correct split detection. This example uses a
four-chassis IRF fabric at the distribution layer. The access-layer and distribution-layer IRF fabrics
are the intermediate device of each other, as shown in Figure 4.
In this example, each member in the access-layer IRF fabric has only two links to the
distribution-layer IRF fabric. For high availability, you can connect each member in the access-layer
IRF fabric to each member in the distribution-layer IRF fabric. After you aggregate these links, they
are regarded as one link in the topology.
For quick split detection, assign high-speed ports for uplink aggregation connections. This example
uses 100-GE ports.
Figure 4 LACP MAD
Network connectivity configuration
In a Layer 2 network, IRF is typically used with link aggregation to simplify the network topology.
For high availability, you can connect each host or server to two ToR switches in the access-layer
IRF fabric, and aggregate the links. On each link aggregation, you do not need to run the spanning
tree protocol feature because an IRF fabric appears as one node in the network.
For VLAN tags to be processed correctly, assign ports and aggregate interfaces to the correct
VLANs.
NOTE:
A link aggregation could span one member device, some of the member devices, or all member
devices, depending on the link redundancy requirements and number of available links. The link
aggregation used for LACP MAD must span all member devices.
IRF
Rack Rack
HGE1/0/29
HGE1/0/30
HGE2/0/29
HGE2/0/30 HGE3/0/29 HGE3/0/30 HGE4/0/29 HGE4/0/30
HGE1/2/0/1 HGE1/3/0/1
HGE2/2/0/1
HGE2/3/0/1
HGE3/2/0/1 HGE3/3/0/1
HGE4/2/0/1 HGE4/3/0/1
1234
12 34

7
Figure 5 Connection diagram for the IRF fabrics in a Layer 2 network
Software versions used
This configuration example was created and verified on R6301.
Restrictions and guidelines
Restrictions and guidelines for LACP MAD configuration
When you configure LACP MAD, follow these restrictions and guidelines:
•You only need to run LACP MAD on a single link aggregation for IRF split detection.
•The link aggregation must use dynamic aggregation mode.
•The link aggregation must have a minimum of one member link from each member device.
Restrictions and guidelines for IRF physical interfaces
Make sure all IRF physical interfaces are operating in bridge mode (the default).
Procedures
This example assumes that the distribution-layer IRF fabric has been set up.
Setting up the access-layer IRF fabric
1. Configure Device A:
# Shut down the physical interfaces used for IRF connection. This example uses the port group
that contains HundredGigE 1/0/25 to HundredGigE 1/0/28 for IRF connection.
<DeviceA> system-view
[DeviceA] interface range hundredgige 1/0/25 to hundredgige 1/0/28
[DeviceA-if-range] shutdown

8
[DeviceA-if-range] quit
# Bind HundredGigE 1/0/25 and HundredGigE 1/0/26 to IRF-port 1/1.
[DeviceA] irf-port 1/1
[DeviceA-irf-port1/1] port group interface hundredgige 1/0/25
[DeviceA-irf-port1/1] port group interface hundredgige 1/0/26
[DeviceA-irf-port1/1] quit
# Bind HundredGigE 1/0/27 and HundredGigE 1/0/28 to IRF-port 1/2.
[DeviceA] irf-port 1/2
[DeviceA-irf-port1/2] port group interface hundredgige 1/0/27
[DeviceA-irf-port1/2] port group interface hundredgige 1/0/28
[DeviceA-irf-port1/2] quit
# Bring up the physical interfaces.
[DeviceA] interface range hundredgige 1/0/25 to hundredgige 1/0/28
[DeviceA-if-range] undo shutdown
[DeviceA-if-range] quit
# Assign IRF member priority 31 to Device A. This priority is high enough to ensure that Device
A can be elected as the master.
[DeviceA] irf member 1 priority 31
# Save the running configuration.
[DeviceA] quit
<DeviceA> save
# Activate the IRF port configuration.
<DeviceA> system-view
[DeviceA] irf-port-configuration active
2. Configure Device B:
# Assign member ID 2 to Device B, and reboot the device to effect the member ID change.
<DeviceB> system-view
[DeviceB] irf member 1 renumber 2
Renumbering the member ID may result in configuration change or loss. Continue? [Y/N]:y
[DeviceB] quit
<DeviceB> reboot
# Shut down the physical interfaces used for IRF connection. This example uses the port group
that contains HundredGigE 2/0/25 to HundredGigE 2/0/28 for IRF connection.
<DeviceB> system-view
[DeviceB] interface range hundredgige 2/0/25 to hundredgige 2/0/28
[DeviceB-if-range] shutdown
[DeviceB-if-range] quit
# Bind HundredGigE 2/0/25 and HundredGigE 2/0/26 to IRF-port 2/1.
[DeviceB] irf-port 2/1
[DeviceB-irf-port2/1] port group interface hundredgige 2/0/25
[DeviceB-irf-port2/1] port group interface hundredgige 2/0/26
[DeviceB-irf-port2/1] quit
# Bind HundredGigE 2/0/27 and HundredGigE 2/0/28 to IRF-port 2/2.
[DeviceB] irf-port 2/2
[DeviceB-irf-port2/2] port group interface hundredgige 2/0/27
[DeviceB-irf-port2/2] port group interface hundredgige 2/0/28
[DeviceB-irf-port2/2] quit
# Bring up the physical interfaces.

9
[DeviceB] interface range hundredgige 2/0/25 to hundredgige 2/0/28
[DeviceB-if-range] undo shutdown
[DeviceB-if-range] quit
# Save the running configuration.
[DeviceB] quit
<DeviceB> save
# Connect Device B to Device A, as shown in Figure 3.
# Activate the IRF port configuration.
<DeviceB> system-view
[DeviceB] irf-port-configuration active
Device B has a lower priority than Device A, and Device B fails master election and reboots. A
two-chassis IRF fabric is formed.
3. Configure Device C:
# Assign member ID 3 to Device C, and reboot the device to effect the member ID change.
<DeviceC> system-view
[DeviceC] irf member 1 renumber 3
Renumbering the member ID may result in configuration change or loss. Continue? [Y/N]:y
[DeviceC] quit
<DeviceC> reboot
# Shut down the physical interfaces used for IRF connection. This example uses the port group
that contains HundredGigE 3/0/25 to HundredGigE 3/0/28 for IRF connection.
<DeviceC> system-view
[DeviceC] interface range hundredgige 3/0/25 to hundredgige 3/0/28
[DeviceC-if-range] shutdown
[DeviceC-if-range] quit
# Bind HundredGigE 3/0/25 and HundredGigE 3/0/26 IRF-port 3/1.
[DeviceC] irf-port 3/1
[DeviceC-irf-port3/1] port group interface hundredgige 3/0/25
[DeviceC-irf-port3/1] port group interface hundredgige 3/0/26
[DeviceC-irf-port3/1] quit
# Bind HundredGigE 3/0/27 and HundredGigE 3/0/28 to IRF-port 3/2.
[DeviceC] irf-port 3/2
[DeviceC-irf-port3/2] port group interface hundredgige 3/0/27
[DeviceC-irf-port3/2] port group interface hundredgige 3/0/28
[DeviceC-irf-port3/2] quit
# Bring up the physical interfaces.
[DeviceC] interface range hundredgige 3/0/25 to hundredgige 3/0/28
[DeviceC-if-range] undo shutdown
[DeviceC-if-range] quit
# Save the running configuration.
[DeviceC] quit
<DeviceC> save
# Connect Device C to Device B, as shown in Figure 3.
# Activate the IRF port configuration.
<DeviceC> system-view
[DeviceC] irf-port-configuration active
Device C reboots to join the IRF fabric. A three-chassis IRF fabric is formed.

10
4. Configure Device D:
# Assign member ID 4 to Device D, and reboot the device to effect the member ID change.
<DeviceD> system-view
[DeviceD] irf member 1 renumber 4
Renumbering the member ID may result in configuration change or loss. Continue? [Y/N]:y
[DeviceD] quit
<DeviceD> reboot
# Shut down the physical interfaces used for IRF connection. This example uses the port group
that contains HundredGigE 4/0/25 to HundredGigE 4/0/28 for IRF connection.
<DeviceD> system-view
[DeviceD] interface range hundredgige 4/0/25 to hundredgige 4/0/28
[DeviceD-if-range] shutdown
[DeviceD-if-range] quit
# Bind HundredGigE 4/0/25 and HundredGigE 4/0/26 to IRF-port 4/1.
[DeviceD] irf-port 4/1
[DeviceD-irf-port4/1] port group interface hundredgige 4/0/25
[DeviceD-irf-port4/1] port group interface hundredgige 4/0/26
[DeviceD-irf-port4/1] quit
# Bind HundredGigE 4/0/27 and HundredGigE 4/0/28 to IRF-port 4/2.
[DeviceD] irf-port 4/2
[DeviceD-irf-port4/2] port group interface hundredgige 4/0/27
[DeviceD-irf-port4/2] port group interface hundredgige 4/0/28
[DeviceD-irf-port4/2] quit
# Bring up the physical interfaces.
[DeviceD] interface range hundredgige 4/0/25 to hundredgige 4/0/28
[DeviceD-if-range] undo shutdown
[DeviceD-if-range] quit
# Save the running configuration.
[DeviceD] quit
<DeviceD> save
# Connect Device D to Device A and Device C, as shown in Figure 3.
# Activate the IRF port configuration.
<DeviceD> system-view
[DeviceD] irf-port-configuration active
Device D reboots to join the IRF fabric. A four-chassis IRF fabric is formed. The command
prompt changes to the name of Device A (the master).
5. Change the name of the IRF fabric.
<DeviceA> system-view
[DeviceA] sysname IRF
[IRF]
Configuring LACP MAD
1. Configure the IRF fabric at the access layer:
# Assign domain ID 1 to the IRF fabric.
<IRF> system-view
[IRF] irf domain 1
# Create Bridge-Aggregation 2, set its aggregation mode to dynamic, and enable LACP MAD.

11
[IRF] interface bridge-aggregation 2
[IRF-Bridge-Aggregation2] link-aggregation mode dynamic
[IRF-Bridge-Aggregation2] mad enable
You need to assign a domain ID (range: 0-4294967295)
[Current domain is: 1]:
The assigned domain ID is: 1
[IRF-Bridge-Aggregation2] quit
# Create a named interface range that contains uplink ports to the distribution layer IRF fabric.
In this example, the interface range name is lacp, and each member device has two links to the
distribution layer IRF fabric.
[IRF] interface range name lacp interface hundredgige 1/0/29 to hundredgige 1/0/30
hundredgige 2/0/29 to hundredgige 2/0/30 hundredgige 3/0/29 to hundredgige 3/0/30
hundredgige 4/0/29 to hundredgige 4/0/30
# Assign the ports in the interface range to Bridge-Aggregation 2.
[IRF-if-range-lacp] port link-aggregation group 2
[IRF-if-range-lacp] quit
2. Configure the IRF fabric at the distribution layer:
# Assign IRF domain ID 2 to the IRF fabric.
<Sysname> system-view
[Sysname] irf domain 2
# Create Bridge-Aggregation 2, set its aggregation mode to dynamic, and enable LACP MAD.
[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: 2]:
The assigned domain ID is: 2
[Sysname-Bridge-Aggregation2] quit
# Assign the downlink ports that connectto the access-layer IRF fabric to Bridge-Aggregation 2.
[Sysname] interface hundredgige 1/2/0/1
[Sysname-HundredGigE1/2/0/1] port link-aggregation group 2
[Sysname-HundredGigE1/2/0/1] quit
[Sysname] interface hundredgige 1/3/0/1
[Sysname-HundredGigE1/3/0/1] port link-aggregation group 2
[Sysname-HundredGigE1/3/0/1] quit
[Sysname] interface hundredgige 2/2/0/1
[Sysname-HundredGigE2/2/0/1] port link-aggregation group 2
[Sysname-HundredGigE2/2/0/1] quit
[Sysname] interface hundredgige 2/3/0/1
[Sysname-HundredGigE2/3/0/1] port link-aggregation group 2
[Sysname-HundredGigE2/3/0/1] quit
[Sysname] interface hundredgige 3/2/0/1
[Sysname-HundredGigE3/2/0/1] port link-aggregation group 2
[Sysname-HundredGigE3/2/0/1] quit
[Sysname] interface hundredgige 3/3/0/1
[Sysname-HundredGigE3/3/0/1] port link-aggregation group 2
[Sysname-HundredGigE3/3/0/1] quit
[Sysname] interface hundredgige 4/2/0/1

12
[Sysname-HundredGigE4/2/0/1] port link-aggregation group 2
[Sysname-HundredGigE4/2/0/1] quit
[Sysname] interface hundredgige 4/3/0/1
[Sysname-HundredGigE4/3/0/1] port link-aggregation group 2
[Sysname-HundredGigE4/3/0/1] quit
Configuring network connectivity settings
This section provides the VLAN and link aggregation configuration procedure for end devices (for
example, a server in VLAN 10).
1. Configure the IRF fabric at the access layer:
# Create VLAN 10.
<IRF> system-view
[IRF] vlan 10
# Create Bridge-Aggregation 3, and set its aggregation mode to dynamic.
[IRF] interface bridge-aggregation 3
[IRF-Bridge-Aggregation3] link-aggregation mode dynamic
[IRF-Bridge-Aggregation3] quit
# Assign Ten-GigabitEthernet 1/0/10 and Ten-GigabitEthernet 2/0/10 to the aggregation group
for Bridge-Aggregation 3.
[IRF] interface ten-gigabitethernet 1/0/10
[IRF-Ten-GigabitEthernet1/0/10] port link-aggregation group 3
[IRF-Ten-GigabitEthernet1/0/10] quit
[IRF] interface ten-gigabitethernet 2/0/10
[IRF-Ten-GigabitEthernet2/0/10] port link-aggregation group 3
[IRF-Ten-GigabitEthernet2/0/10] quit
# Assign Bridge-Aggregation 3 to VLAN 10.
[IRF] interface bridge-aggregation 3
[IRF-Bridge-Aggregation3] port access vlan 10
[IRF-Bridge-Aggregation3] quit
# Configure Bridge-Aggregation 2 as a trunk port, and assign it to VLAN 10.
[IRF] interface bridge-aggregation 2
[IRF-Bridge-Aggregation2] port link-type trunk
[IRF-Bridge-Aggregation2] port trunk permit vlan 10
[IRF-Bridge-Aggregation2] quit
2. Configure the IRF fabric at the distribution layer:
# Create VLAN 10.
<Sysname> system-view
[Sysname] vlan 10
# Configure Bridge-Aggregation 2 as a trunk port, and assign it to VLAN 10.
[Sysname] interface bridge-aggregation 2
[Sysname-Bridge-Aggregation2] port link-type trunk
[Sysname-Bridge-Aggregation2] port trunk permit vlan 10
[Sysname-Bridge-Aggregation2] quit

13
Verifying the configuration
Verify the IRF setup, multichassis link aggregations, ring topology, and LACP MAD.
Verifying the IRF setup
# Verify that the IRF fabric has been formed.
[IRF] display irf
MemberID Role Priority CPU-Mac Description
*+1 Master 31 00e0-fc0f-8c02 ---
2 Standby 1 00e0-fc0f-8c03 ---
3 Standby 1 00e0-fc0f-8c04 ---
4 Standby 1 00e0-fc0f-8c05 ---
--------------------------------------------------
* indicates the device is the master.
+ indicates the device through which the user logs in.
The Bridge MAC of the IRF is: 0cda-414a-859b
Auto upgrade : yes
Mac persistent : 12 min
Domain ID : 1
The output shows that the IRF fabric has four member devices.
# Verify IRF fabric connectivity.
[IRF] display irf topology
Topology Info
-------------------------------------------------------------------------
IRF-Port1 IRF-Port2
MemberID Link neighbor Link neighbor Belong To
1 UP 2 UP 4 0cda-414a-859b
2 UP 3 UP 1 0cda-414a-859b
3 UP 4 UP 2 0cda-414a-859b
4 UP 1 UP 3 0cda-414a-859b
The output shows that all the IRF links are in UP state. The four-chassis IRF fabric is established.
Verifying the link backup function of multichassis
aggregations
# Ping the IP address of the IRF fabric at the distribution layer from a server.
C:\Users>ping 10.153.116.111 –t
# On the IRF fabric at the access layer, shut down Ten-GigabitEthernet 1/0/10, a member port of
Bridge-Aggregation 3.
[IRF] interface ten-gigabitethernet 1/0/10
[IRF-Ten-GigabitEthernet1/0/10] shutdown
[IRF-Ten-GigabitEthernet1/0/10] quit
# Observe the output on the server configuration terminal.
Pinging 10.153.116.111 with 32 bytes of data:

14
Reply from 10.153.116.111: bytes=32 time<1ms TTL=128
Reply from 10.153.116.111: bytes=32 time<1ms TTL=128
Request timed out.
Reply from 10.153.116.111: bytes=32 time<1ms TTL=128
Reply from 10.153.116.111: bytes=32 time<1ms TTL=128
The output shows that the IP address can be pinged after transient traffic disruption.
# On the IRF fabric at the access layer, shut down HundredGigE 1/0/29 and HundredGigE 1/0/30
(member ports of Bridge-Aggregation 2). The server cannot access the distribution layer through
Device A.
[IRF] interface range hundredgige 1/0/29 hundredgige 1/0/30
[IRF-if-range] shutdown
[IRF-if-range] quit
# Observe the output on the server configuration terminal.
Pinging 10.153.116.111 with 32 bytes of data:
Reply from 10.153.116.111: bytes=32 time<1ms TTL=128
Reply from 10.153.116.111: bytes=32 time<1ms TTL=128
Request timed out.
Reply from 10.153.116.111: bytes=32 time<1ms TTL=128
Reply from 10.153.116.111: bytes=32 time<1ms TTL=128
The output shows that the IP address can be pinged after transient traffic disruption. The server
accesses the distribution layer through another device in the access-layer IRF fabric.
Verifying link failure protection of the ring topology
# Disconnect all IRF links between two IRF member devices. (Details not shown.)
# Verify that the IRF fabric operates correctly in daisy-chain topology. (Details not shown.)
Verifying the LACP MAD configuration
# As shown in Figure 6, disconnect two IRF connections: one between Device Aand Device D, and
the other between Device B and Device C. The disconnect actions cause the IRF fabric to break up
into two parts.
Figure 6 IRF split
# Observe the output messages to verify that LACP MAD takes the following actions:
•Changes IRF 2 (Device C and Device D) to the Recovery state, because the master device in
IRF 1 has a lower member ID than the master device in IRF 2.
•Shuts down all service interfaces on Device C and Device D, except for the IRF physical
interfaces and the service interfaces excluded from the MAD shutdown action.

15
The following is the sample output on IRF 1:
%Jan 1 05:19:10:176 2019 H3C STM/3/STM_LINK_STATUS_DOWN: IRF port 2 is down.
%Jan 1 05:19:10:184 2019 H3C IFNET/3/PHY_UPDOWN: HundredGigE1/0/27 link status is down.
%Jan 1 05:19:10:184 2019 H3C IFNET/3/PHY_UPDOWN: HundredGigE1/0/28 link status is down.
%Jan 1 05:19:10:176 2019 H3C STM/3/STM_LINK_STATUS_DOWN: IRF port 1 is down.
%Jan 1 05:19:10:184 2019 H3C IFNET/3/PHY_UPDOWN: HundredGigE2/0/25 link status is down.
%Jan 1 05:19:10:184 2019 H3C IFNET/3/PHY_UPDOWN: HundredGigE2/0/26 link status is down.
%Jan 1 05:19:10:186 2019 H3C DEV/3/BOARD_REMOVED: Board is removed from Slot 3, type is
MAIN_BOARD_TYPE_54QT.
%Jan 1 05:19:10:186 2019 H3C DEV/3/BOARD_REMOVED: Board is removed from Slot 4, type is
MAIN_BOARD_TYPE_54QT.
The following is the sample output from IRF 2:
%Jan 1 05:53:20:784 2019 H3C HA/5/HA_STANDBY_TO_MASTER: Standby board in slot 3 changes
to master.
%Jan 1 05:53:20:831 2019 H3C DEV/3/BOARD_REMOVED: Board is removed from Slot 1, type is
MAIN_BOARD_TYPE_56HF.
%Jan 1 05:53:20:831 2019 H3C DEV/3/BOARD_REMOVED: Board is removed from Slot 2, type is
MAIN_BOARD_TYPE_56HF.
%Jan 1 05:53:20:860 2019 H3C DEV/1/MAD_DETECT: Multi-active devices detected, please fix
it.
%Jan 1 05:53:20:886 2019 H3C IFNET/3/PHY_UPDOWN: M-GigabitEthernet0/0/0 link status is
down.
%Jan 1 05:53:20:887 2019 H3C IFNET/5/LINK_UPDOWN: Line protocol on the interface
M-GigabitEthernet0/0/0 is down.
%Jan 1 05:53:20:912 2019 H3C IFNET/3/PHY_UPDOWN: HundredGigE3/0/29 link status is down.
%Jan 1 05:53:20:914 2019 H3C IFNET/5/LINK_UPDOWN: Line protocol on the interface
HundredGigE3/0/29 is down.
%Jan 1 05:53:20:912 2019 H3C IFNET/3/PHY_UPDOWN: HundredGigE3/0/30 link status is down.
%Jan 1 05:53:20:914 2019 H3C IFNET/5/LINK_UPDOWN: Line protocol on the interface
HundredGigE3/0/30 is down.
%Jan 1 05:53:20:912 2019 H3C IFNET/3/PHY_UPDOWN: HundredGigE4/0/29 link status is down.
%Jan 1 05:53:20:914 2019 H3C IFNET/5/LINK_UPDOWN: Line protocol on the interface
HundredGigE4/0/29 is down.
%Jan 1 05:53:20:912 2019 H3C IFNET/3/PHY_UPDOWN: HundredGigE4/0/30 link status is down.
%Jan 1 05:53:20:914 2019 H3C IFNET/5/LINK_UPDOWN: Line protocol on the interface
HundredGigE4/0/30 is down.
The output shows that initially IRF 2 took over the master role because it falsely identified that IRF 1
has failed. LACP MAD shut down the service interfaces on IRF 2 after it detected an MAD conflict.
# If IRF 1 fails, use the mad restore command on IRF 2 to recover the member devices and bring
up all service interfaces that have been shut down by LACP MAD.
NOTE:
You can log in to IRF 2 through the console port on Device C or Device D. If you have excluded a
service interface from the MAD shutdown action, you can Telnet to IRF 2 through the service
interface.
<Sysname> system-view
[Sysname] mad restore
This command will restore the device from multi-active conflict state. Continue? [Y/N]:y
Restoring from multi-active conflict state, please wait...

16
[Sysname]
%Jan 1 05:24:41:249 2019 H3C IFNET/3/PHY_UPDOWN: Ten-GigabitEthernet2/0/10 link status
is up.
%Jan 1 05:24:41:249 2019 H3C IFNET/5/LINK_UPDOWN: Line protocol on the interface
Ten-GigabitEthernet2/0/10 is up.
%Jan 1 05:24:41:325 2019 H3C IFNET/3/PHY_UPDOWN: HundredGigE2/0/29 link status is up.
%Jan 1 05:24:41:325 2019 H3C IFNET/3/PHY_UPDOWN: HundredGigE2/0/30 link status is up.
%Jan 1 05:24:46:266 2019 H3C IFNET/3/PHY_UPDOWN: M-GigabitEthernet0/0/0 link status is
up.
%Jan 1 05:24:46:268 2019 H3C IFNET/5/LINK_UPDOWN: Line protocol on the interface
M-GigabitEthernet0/0/0 is up.
The output shows that network connectivity of IRF 2 has been restored.
# Remove all IRF 1 and IRF link failures.
# After the link failure of IRF 1 is removed, the IRF ports recover. The following is the sample output
on IRF 1:
%Jan 1 05:29:06:913 2019 H3C IFNET/3/PHY_UPDOWN: HundredGigE1/0/27 link status is up.
%Jan 1 05:29:06:914 2019 H3C IFNET/5/LINK_UPDOWN: Line protocol on the interface
HundredGigE1/0/27 is up.
%Jan 1 05:29:06:913 2019 H3C IFNET/3/PHY_UPDOWN: HundredGigE1/0/28 link status is up.
%Jan 1 05:29:06:914 2019 H3C IFNET/5/LINK_UPDOWN: Line protocol on the interface
HundredGigE1/0/28 is up.
%Jan 1 05:29:07:106 2019 H3C STM/6/STM_LINK_STATUS_UP: IRF port 2 is up.
%Jan 1 05:29:07:810 2019 H3C STM/4/STM_LINK_RECOVERY: Merge occurs.
IRF 2 reboots automatically to merge with IRF 1.
# Verify that the IRF fabric is recovered with Device A as the master.
<IRF> display irf
MemberID Role Priority CPU-Mac Description
*+1 Master 31 00e0-fc0f-8c02 ---
2 Standby 1 00e0-fc0f-8c03 ---
3 Standby 1 00e0-fc0f-8c04 ---
4 Standby 1 00e0-fc0f-8c05 ---
--------------------------------------------------
* indicates the device is the master.
+ indicates the device through which the user logs in.
The Bridge MAC of the IRF is: 0cda-414a-859b
Auto upgrade : yes
Mac persistent : 12 min
Domain ID : 1
Configuration files
•IRF fabric at the access layer:
#
sysname IRF
#
irf domain 1
irf member 1 priority 31

17
#
vlan 10
#
irf-port1/1
port group interface HundredGigE1/0/25
port group interface HundredGigE1/0/26
#
irf-port1/2
port group interface HundredGigE1/0/27
port group interface HundredGigE1/0/28
#
irf-port2/1
port group interface HundredGigE2/0/25
port group interface HundredGigE2/0/26
#
irf-port2/2
port group interface HundredGigE2/0/27
port group interface HundredGigE2/0/28
#
irf-port3/1
port group interface HundredGigE3/0/25
port group interface HundredGigE3/0/26
#
irf-port3/2
port group interface HundredGigE3/0/27
port group interface HundredGigE3/0/28
#
irf-port4/1
port group interface HundredGigE4/0/25
port group interface HundredGigE4/0/26
#
irf-port4/2
port group interface HundredGigE4/0/27
port group interface HundredGigE4/0/28
#
interface Bridge-Aggregation2
port link-type trunk
port trunk permit vlan 1 10
link-aggregation mode dynamic
mad enable
#
interface Bridge-Aggregation3
port access vlan 10
link-aggregation mode dynamic
#
interface HundredGigE1/0/29
port link-mode bridge
port link-type trunk

18
port trunk permit vlan 1 10
port link-aggregation group 2
#
interface HundredGigE1/0/30
port link-mode bridge
port link-type trunk
port trunk permit vlan 1 10
port link-aggregation group 2
#
interface HundredGigE2/0/29
port link-mode bridge
port link-type trunk
port trunk permit vlan 1 10
port link-aggregation group 2
#
interface HundredGigE2/0/30
port link-mode bridge
port link-type trunk
port trunk permit vlan 1 10
port link-aggregation group 2
#
interface HundredGigE3/0/29
port link-mode bridge
port link-type trunk
port trunk permit vlan 1 10
port link-aggregation group 2
#
interface HundredGigE3/0/30
port link-mode bridge
port link-type trunk
port trunk permit vlan 1 10
port link-aggregation group 2
#
interface HundredGigE4/0/29
port link-mode bridge
port link-type trunk
port trunk permit vlan 1 10
port link-aggregation group 2
#
interface HundredGigE4/0/30
port link-mode bridge
port link-type trunk
port trunk permit vlan 1 10
port link-aggregation group 2
#
interface Ten-GigabitEthernet1/0/10
port link-mode bridge
port access vlan 10