Description

This article describes a controlled procedure for adding an SPC II card to an SRX5000 Series chassis cluster while minimizing service impact. The process uses node isolation, controlled failover, and staged recovery to safely upgrade both cluster members.

Symptoms

This is an informational KB.

Use this procedure when:

  • An additional SPC II card needs to be installed in an SRX5k chassis cluster
  • The upgrade must be completed with controlled failover and minimal service risk
  • A staged node-by-node hardware insertion is preferred

Solution

  • Perform the upgrade during a planned maintenance window, following the sequence below.

    1) Prepare the cluster

    • Fail over all redundancy groups to Node 0
    • Disable the physical transit interfaces on Node 1
    • Disable preempt for RG1+
    • Disable interface monitoring and IP monitoring
    • Disable TCP SYN check
    • Disable TCP sequence check

    2) Isolate the secondary node

    • Move the control link to unused or non-existent ports to simulate control link failure
    • Remove the fabric port configuration to simulate fabric link failure
    • Commit the configuration
    • Confirm that both nodes are isolated before proceeding

    3) Upgrade Node 1

    • Power down Node 1
    • Install the additional SPC II card
    • Power on Node 1
    • Perform a health check and confirm the node is operating normally

    4) Shift traffic to Node 1

    • Disable the physical revenue interfaces on Node 0
    • Enable the physical revenue interfaces on Node 1
    • Verify that traffic is successfully forwarded through Node 1

    5) Upgrade Node 0

    • Power down Node 0
    • Install the additional SPC II card
    • Power on Node 0
    • Perform a health check and confirm the node is operating normally

    6) Restore cluster connectivity

    • Reconfigure the control and fabric ports on Node 0, then halt the node
    • Reconfigure the control and fabric ports on Node 1
    • Reboot Node 0

    7) Restore normal operation

    • Re-enable the physical revenue interfaces on Node 0
    • Verify overall chassis cluster health
    • Re-enable:
      • TCP SYN check
      • TCP sequence check
      • Preempt for RG1+
      • Interface/IP monitoring

    After both SPC II cards are installed and the cluster is restored, confirm the following:

    • Both nodes are online and healthy
    • Control and fabric links are restored
    • Revenue traffic is forwarding normally
    • Redundancy groups are in the expected state
    • Cluster protection and monitoring features are re-enabled

    To safely add an SPC II card to an SRX5k chassis cluster, isolate the nodes, upgrade one node at a time, validate traffic handling after each step, and then restore the cluster to its normal state.

Modification History

2024-09-06 : Article Created