Networking Requirement

An enterprise builds a data center. The data center network must provide high reliability on the core layer and have a simple structure to facilitate configuration and management.

To meet requirements of the enterprise, core switches SwitchA and SwitchB set up a stack. Figure 4-1 shows the network topology.

Figure 4-1 Stack networking

Configuration Roadmap

The configuration roadmap is as follows:

1. Make a stack plan.
2. Connect SwitchA and SwitchB using stack cables based on the stack plan.
3. Configure stack software on SwitchA and SwitchB, including the stack member ID, stack priority, stack domain ID, stack connection mode, and stack port.
4. Save the configurations of SwitchA and SwitchB and enable the stacking function on the two switches.
5. Check whether a stack has been set up successfully.
6. Configure DAD.

Procedure

1. Make a stack plan.

   Table 4-1 shows a stack plan.

   Table 4-1 Stack plan

   Item	SwitchA	SwitchB
   Stack member role	Master	Standby
   Stack member ID	1	2
   Stack priority	150	100
   Stack domain ID	10	10
   Stack connection mode	MPU connection mode	MPU connection mode
   Ports used for stack connections	40GE1/0/1-40GE1/0/2 40GE2/0/1-40GE2/0/2	40GE1/0/1-40GE1/0/2 40GE2/0/1-40GE2/0/2
   DAD port	10GE3/0/5 10GE4/0/5	10GE3/0/5 10GE4/0/5

2. Connect stack cables.

   In Figure 4-2, SwitchA and SwitchB are connected using stack cables based on the stack plan. The two switches are connected using SIP ports on MPUs, stack member ports on LPUs, and DAD ports.

   Figure 4-2 Stack connections
   

3. Configure stack attributes.

   # On SwitchA, set the stack priority to 150 and stack domain ID to 10. In this example, SwitchA retains the default stack member ID 1 and default stack connection mode MPU connection, and you do not need to configure the two parameters.

   <HUAWEI> system-view 
   [~HUAWEI] sysname SwitchA 
   [*HUAWEI] commit 
   [~SwitchA] stack 
   [~SwitchA-stack] stack priority 150 
   [*SwitchA-stack] stack domain 10 
   [*SwitchA-stack] quit 
   [*SwitchA] commit

   # On SwitchB, set the stack member ID to 2 and stack domain ID to 10. In this example, SwitchB retains the default stack priority 100 and default stack connection mode MPU connection, and you do not need to configure the two parameters.

   <HUAWEI> system-view 
   [~HUAWEI] sysname SwitchB 
   [*HUAWEI] commit 
   [~SwitchB] stack 
   [~SwitchB-stack] stack member 2 
   Warning: The device will use the configuration of member ID 2 after the device resets. Continue? [Y/N]: y 
   [*SwitchB-stack] stack domain 10 
   [*SwitchB-stack] quit 
   [*SwitchB] commit

4. Configure logical stack ports.

   # Create a logical stack port on SwitchA, and add ports 40GE1/0/1 to 40GE1/0/2 and 40GE2/0/1 to 40GE2/0/2 to the logical stack port. # Create a logical stack port on SwitchB, and add ports 40GE1/0/1 to 40GE1/0/2 and 40GE2/0/1 to 40GE2/0/2 to the logical stack port. The configuration of SwitchB is similar to that of SwitchA, and is not mentioned here.

   [~SwitchA] port-group group1 
   [*SwitchA-port-group-group1] group-member 40ge 1/0/1 to 40ge 1/0/2 
   [*SwitchA-port-group-group1] group-member 40ge 2/0/1 to 40ge 2/0/2 
   [*SwitchA-port-group-group1] shutdown 
   [*SwitchA-port-group-group1] quit 
   [*SwitchA] commit 
   [~SwitchA] interface stack-port 1 
   [*SwitchA-Stack-Port1] port member-group interface 40ge 1/0/1 to 1/0/2 
   Warning: After the configuration is complete, 
   1.The interface(s) (40GE1/0/1-1/0/2) will be converted to stack mode and be configured with the port crc-statistics trigger error-down command if the configuration does not exist. 
   2.The interface(s) may go Error-Down (crc-statistics) because there is no shutdown configuration on the interfaces.Continue? [Y/N]: y 
   [*SwitchA-Stack-Port1] port member-group interface 40ge 2/0/1 to 2/0/2 
   Warning: After the configuration is complete, 
   1.The interface(s) (40GE2/0/1-2/0/2) will be converted to stack mode and be configured with the port crc-statistics trigger error-down command if the configuration does not exist. 
   2.The interface(s) may go Error-Down (crc-statistics) because there is no shutdown configuration on the interfaces.Continue? [Y/N]: y 
   [*SwitchA-Stack-Port1] quit 
   [*SwitchA] commit

   After a port is configured as a stack member port, the port crc-statistics trigger error-down command is automatically configured on this port.

   # After physical ports on SwitchA and SwitchB are added to a logical stack port, enable the ports that are shut down. The configuration of SwitchB is similar to that of SwitchA, and is not mentioned here.

   [~SwitchA] port-group group1 
   [~SwitchA-port-group-group1] undo shutdown 
   [*SwitchA-port-group-group1] quit 
   [*SwitchA] commit 
   [~SwitchA] quit

   Shutting down and then enabling ports prevent these ports from generating CRC error packets and entering the Error-Down state. When ports are Up, CRC error packets may be generated on these ports if ports on both ends of a link are a stack member port and an ordinary service port respectively.

5. Check the stack configuration.

   # After the preceding configuration is complete, run the display stack configuration command to check whether the configuration is consistent with the stack plan. If not, modify the configuration. The command output on SwitchA is used as an example.

   <SwitchA> display stack configuration 
   Oper          : Operation 
   Conf          : Configuration 
   LC            : LineCard 
   MB            : MainBoard 
   *             : Offline configuration 
   Isolated Port : The port is in stack mode, but does not belong to any Stack-Port 
    
   Attribute Configuration: 
   --------------------------------------------------------------- 
    MemberID      Domain         Priority       Mode     Enable 
   Oper(Conf) Oper(Conf)      Oper(Conf) Oper(Conf) Oper 
   --------------------------------------------------------------- 
   1(1)         --(10)          100(150)     MB(MB)       Disable 
   --------------------------------------------------------------- 
    
   Stack-Port Configuration: 
   -------------------------------------------------------------------------------- 
   Stack-Port      Member Ports 
   -------------------------------------------------------------------------------- 
   Stack-Port1     40GE1/0/1           40GE1/0/2           40GE2/0/1 
                   40GE2/0/2 
   --------------------------------------------------------------------------------

6. Save the configuration and enable the stacking function.

   # Save the configuration of SwitchA and enable the stacking function. You are advised to enable the stacking function first on SwitchA that is specified as the master switch. This ensures that SwitchA restarts first and becomes the master switch.

   <SwitchA> save 
   Warning: The current configuration will be written to the device. Continue? [Y/N]: y 
   <SwitchA> system-view 
   [~SwitchA] stack 
   [~SwitchA-stack] stack enable 
   Warning: Make sure that one or more dual-active detection methods are configured once the conversion is complete and the device enters the stack mode. Switches working in different forward modes cannot set up a CSS. 
   Current configuration will be converted to the next startup saved-configuration file of stack mode. 
   System will reboot. Continue? [Y/N]: y

   After the stacking function is enabled on a switch, this switch automatically restarts.

   # Save the configuration of SwitchB and enable the stacking function.

   <SwitchB> save 
   Warning: The current configuration will be written to the device. Continue? [Y/N]: y 
   <SwitchB> system-view 
   [~SwitchB] stack 
   [~SwitchB-stack] stack enable 
   Warning: Make sure that one or more dual-active detection methods are configured once the conversion is complete and the device enters the stack mode. Switches working in different forward modes cannot set up a CSS. 
   Current configuration will be converted to the next startup saved-configuration file of stack mode. 
   System will reboot. Continue? [Y/N]: y

7. Check whether a stack has been set up successfully.

   # Log in to the stack through the console port or management port of any switch and run the display stack command to check whether the stack has been set up successfully. To log in the stack through the management port, specify the IP address of the master switch.

   <SwitchA> display stack 
   -------------------------------------------------------------------------------- 
   MemberID Role     MAC              Priority DeviceType       Description 
   -------------------------------------------------------------------------------- 
   1        Master 006d-8835-2b00 150        CE12804 
   2        Standby  006d-8835-2c00 100        CE12804 
   --------------------------------------------------------------------------------

   The preceding command output shows that a stack has been set up successfully and the master switch is the switch with the stack member ID 1, namely, SwitchA.

8. Configure DAD.

   # Configure DAD in direct mode on 10GE1/3/0/5, 10GE1/4/0/5, 10GE2/3/0/5, and 10GE2/4/0/5.

   <SwitchA> system-view 
   [~SwitchA] interface 10ge 1/3/0/5 
   [~SwitchA-10GE1/3/0/5] dual-active detect mode direct 
   Warning: The interface will block common data packets, except BPDU packets. Continue? [Y/N]: y 
   [*SwitchA-10GE1/3/0/5] quit 
   [*SwitchA] interface 10ge 1/4/0/5 
   [*SwitchA-10GE1/4/0/5] dual-active detect mode direct 
   Warning: The interface will block common data packets, except BPDU packets. Continue? [Y/N]: y 
   [*SwitchA-10GE1/4/0/5] quit 
   [*SwitchA] interface 10ge 2/3/0/5 
   [*SwitchA-10GE2/3/0/5] dual-active detect mode direct 
   Warning: The interface will block common data packets, except BPDU packets. Continue? [Y/N]: y 
   [*SwitchA-10GE2/3/0/5] quit 
   [*SwitchA] interface 10ge 2/4/0/5 
   [*SwitchA-10GE2/4/0/5] dual-active detect mode direct 
   Warning: The interface will block common data packets, except BPDU packets. Continue? [Y/N]: y 
   [*SwitchA-10GE2/4/0/5] commit 
   [~SwitchA-10GE2/4/0/5] quit

9. Configure stack member ports to report the Down state after a delay.

   You are advised to set the delay in reporting a port Down event to prevent unstable stack status caused by intermittent port disconnection.

   [~SwitchA] port-group group-member 40ge 1/1/0/1 40ge 1/1/0/2 40ge 1/2/0/1 40ge 1/2/0/2 40ge 2/1/0/1 40ge 2/1/0/2 40ge 2/2/0/1 40ge 2/2/0/2 
   [~SwitchA-port-group] carrier down-hold-time 2000 
   [*SwitchA-port-group] commit 
   [~SwitchA-port-group] return

10. Save the system configuration.

    # After the preceding configuration is complete, run the save command to save the stack configuration immediately.

    <SwitchA> save 
    Warning: The current configuration will be written to the device. Continue? [Y/N]: y

Configuration Files

Stack configuration file

# 
sysname SwitchA 
# 
stack 
 # 
 stack mode 
 # 
 stack member 1 domain 10 
 stack member 1 priority 150 
 # 
 stack member 2 domain 10 
# 
interface Stack-Port1/1 
# 
interface Stack-Port2/1 
# 
interface 10GE1/3/0/5 
 dual-active detect mode direct 
# 
interface 10GE1/4/0/5 
 dual-active detect mode direct 
# 
interface 10GE2/3/0/5 
 dual-active detect mode direct 
# 
interface 10GE2/4/0/5 
 dual-active detect mode direct 
# 
interface 40GE1/1/0/1 
 port mode stack 
 stack-port 1/1 
 port crc-statistics trigger error-down 
 carrier down-hold-time 2000 
# 
interface 40GE1/1/0/2 
 port mode stack 
 stack-port 1/1 
 port crc-statistics trigger error-down 
 carrier down-hold-time 2000 
# 
interface 40GE1/2/0/1 
 port mode stack 
 stack-port 1/1 
 port crc-statistics trigger error-down 
 carrier down-hold-time 2000 
# 
interface 40GE1/2/0/2 
 port mode stack 
 stack-port 1/1 
 port crc-statistics trigger error-down 
 carrier down-hold-time 2000 
# 
interface 40GE2/1/0/1 
 port mode stack 
 stack-port 2/1 
 port crc-statistics trigger error-down 
# 
interface 40GE2/1/0/2 
 port mode stack 
 stack-port 2/1 
 port crc-statistics trigger error-down 
 carrier down-hold-time 2000 
# 
interface 40GE2/2/0/1 
 port mode stack 
 stack-port 2/1 
 port crc-statistics trigger error-down 
 carrier down-hold-time 2000 
# 
interface 40GE2/2/0/2 
 port mode stack 
 stack-port 2/1 
 port crc-statistics trigger error-down 
 carrier down-hold-time 2000 
# 
port-group group1 
 group-member 40GE1/1/0/1 
 group-member 40GE1/1/0/2 
 group-member 40GE1/2/0/1 
 group-member 40GE1/2/0/2 
# 
return