Forum Discussion

iRule's avatar
iRule
Icon for Cirrus rankCirrus
Oct 23, 2020

Impact of forcing the mcpd process to reload the BIG-IP configuration

Dear Community,

 

I am facing HA synchronization failure on F5 vCMP guest HA pair due to difference between bigip.conf file and MCP database in Standby unit. Manual synchronization is configured in HA pair.

I am going to force the mcpd process to reload the BIG-IP configuration according to instruction 1, 2 & 3 in KB https://support.f5.com/csp/article/K13030 on standby unit.

 

Please inform by performing step 1, 2 & 3 from the KB https://support.f5.com/csp/article/K13030 on the standby unit, what will be impact on active unit. Will the production traffic on active unit be interrupted or not.

 

Warm Regards

1 Reply

  • To me mcpd forceload would be the last resort.

    1. Iff the sync is not happening, have you first listed out the status on why sync not happening in 1st place. Run the show cm sync status command and see if there's any conflicts.

    2. Check ntp ​config and time difference if any

    3. You can also run run cm watch device group to see if the commit Id matches, because there are few versions having bug, where config is already happened, but devices show not in sync.

    4. After running save sys config on both boxes, diff the config files.

    5. You can try setting current active as leader on sync group, and then do a full config overwrite on standby box.

    These are few things id try before doing a mcpd forceload.

    ​Coming back to your question, you can set current active box as force active and then perform your change on standby box. There should be NO interruption on the traffic. Make sure you remove force active after your activity.