Description

This article describes the procedure for replacing the Switch Control Board (SCB) or Routing Engine (RE) on a high-end SRX Series device (SRX1400, SRX3400, SRX3600, SRX5400, SRX5600, or SRX5800) that is configured in a chassis cluster.

Symptoms

When replacing most parts on a high-end chassis cluster, the cluster information is not lost and the device is able to rejoin the cluster after reboot. However, when the below parts are replaced, the cluster information is lost.

Parts affected:

  • Switch Control Board (SCB): SRX5400, SRX5600, and SRX5800  
  • Routing Engine (RE): SRX1400, SRX3400, and SRX3600

Related links:

Solution

Follow the instructions below when replacing an RE or SCB in a high-end SRX chassis cluster.

RE Replacement for SRX1400, SRX3400, and SRX3600

Prerequisites:

  1. Disable preempt for redundancy groups.

    deactivate chassis cluster redundancy-group <#> preempt
  2. Obtain the Junos OS software version that is currently in use from the Juniper Download Site .

  3. Save the configuration from the active node.

Steps:

  1. Power down the SRX to have the RE replaced.

  2. Disconnect the control, fabric, and traffic-based interfaces on the powered-down node only.

  3. Replace the RE. You can find detailed instructions on how to replace the RE in the Hardware Guide .

  4. Power on the SRX.

  5. Check the Junos OS version on the active node of the cluster and upgrade or downgrade the Junos OS on the new device so that both nodes are using the same version. For more information, see KB16652 - SRX Getting Started [juniper.net] .

  6. Set the chassis cluster settings. Enter cluster ID information by using the following commands: 
     >set chassis cluster cluster-id X node Y reboot

    Where X is the cluster ID and Y is the node ID.

    After this command is entered, the device will reboot but will not join the cluster.

  7. Upload the saved configuration file to the replacement device in the /var/tmp directory.

  8. Load the configuration from the saved file.

    Enter configuration mode and load the configuration from the file, which was copied into the /var/tmp directory in the preceding step. Use the following command to load the configuration:
    #load override /var/tmp/<filename>
#commit
  9. Power down the device.  
    >request system poweroff
  10. Plug in the control, fabric, and traffic-based interfaces.

  11. Power the device back up.

Post replacement

  1. Re-enable preempt for the redundancy groups (if it was disabled as part of the replacement process). 
    activate chassis cluster redundancy-group <#> preempt
  2. If using PKI Certificates, refer to KB26879 - How to copy certificates after RE replacement in HA cluster [juniper.net] for instructions on copying certificate information to the replacement RE.

SCB Replacement for SRX5800, SRX5600, and SRX5400

Note: If the configuration consists of reth and non-reth interfaces, it is recommended to disconnect the network cables while booting the device for the first time after replacement (of the SCB or RE).

  1. Before you power down the node, first deactivate the fab interfaces within the configuration. This can be run from either node. 
    user@host# deactivate interfaces fab0
user@host# deactivate interfaces fab1
user@host# commit

    Important: Failure to do this first step may result in chassis cluster challenges later on.

  2. Power down the SRX.

  3. Replace the SCB. You can find detailed instructions on how to replace the SCB in the Hardware Guide .

  4. Power on the SRX. At the Junos OS prompt, put the node back into the cluster by using the following command. Make sure that you are in operational mode. You can tell this by the prompt containing a greater than (>) symbol. Here, we are telling the node to reboot as well: 
    user@host> set chassis cluster cluster-id X node Y reboot

    (where x is the cluster ID and Y is the node ID. Make sure that you have console access to the node where the SCB is replaced to do this step.)

  5. When the node is back in the cluster, activate the fab interfaces that you disabled earlier:
user@host# activate interfaces fab0
user@host# activate interfaces fab1 user@host# commit

The cluster should now be operational on both nodes.

Hardware Guide References

Refer to the following documents for hardware replacement procedures:

Modification History

2020-06-23: Article reviewed for accuracy; no changes required; article still valid and relevant

Related Information