How to prepare a standalone SRX Branch to join another SRX Branch configured for chassis cluster using USB snapshot

How to prepare a standalone SRX Branch to join another SRX Branch, which is configured for the chassis cluster.

1. If you are replacing a SRX in the cluster, power down the node to be removed.

2. Remove

from the production node for all redundancy groups and commit.

  • Delete the chassis cluster redundancy-group 1 preempt.

3. Configure new

interfaces on the production node (if using affected versions).

4. Commit and quit.

5. Prepare the new node for isolated testing (no cable connections).

6. Plug in all the modules in the same slots as the production node.

7. Power up the new/replacement node.

8. Take a snapshot of the production node:

  • Insert the USB in the RE module .
  • Execute the
    command.
  • When the above command completes, remove the USB memory stick from the production node.

9. Enable the cluster on the replacement node and boot from the USB snapshot:

  • Execute the
    command (replace X and Y with appropriate values).
  • You will get a notification to reboot; do not reboot now.
  • Insert the USB in the RE module .
  • Execute the
    command, to boot system up with the USB snapshot.

10. After the system boots up from the USB, restore the snapshot to the node’s internal storage:

  • Execute the request system snapshot media internal partition command.

11. After the snapshot completes, reboot the system from the internal storage:

  • Execute the
    command, to boot the system up from the internal flash.
  • After the system returns completely online, remove the USB memory stick.

12. Perform a health check on the replacement system, when it is still unplugged from the network:

  • Does
    show all the modules online? (do not proceed until all are show online; this will take 5-10 minutes).
  • Does
    show no major alarms?
  • Does
    show the primary state for this node?
  • Does
    show no errors for the fabric links?

13. The on-site engineer re-introduces the node to the cluster:

  • Power down the replacement node.
  • Completely cable the replacement node to the production network, including control/fab links.
  • Boot up the replacement node.

14. Perform a health check on the replacement system, after it boots up fully cabled to the network:

  • Does
    show all the modules online? (Do not proceed until all are shown online; this will take 5-10 minutes).
  • Does
    show no major alarms?
  • Does
    show the primary and secondary states, with non-0 priority for both nodes on all redundancy groups?
  • Does
    show no errors for the fabric links?

About the author

Prasanna

Leave a Comment