FTD HA Cluster Only Partially Upgraded

[mumbles202]

Working on a pair of 2130s managed by an FMC that were running 6.2.3.12 and went to upgrade them to 6.6.4 last night.  Did a push first, selected the cluster and confirmed both showed as complete.  Started the install and after completion of the task I have this message:

Update complete, not connected to peer (unable to get status from device)

Connecting to the FTD if I go into system support diagnostic-cli you get the warning that the HA peer has a different version of code, almost lost an ASA HA cluster when only the standby unit has been upgraded.

When I go into /ngfw/var/log/sf/ and look for the 6.6.4 directory to get the logs it doesn't exist.  If I go to /ngfw/var/sf/updates on the FTD that didn't upgrade the file is there.

The cluster had been working fine for years and had no pending deployments/tasks prior to launching the upgrade.  

[Marvin Rhoads]

That's quite odd and obviously not expected behavior.

If it were me, I'd take that issue straight to Cisco TAC.

[mumbles202]

Thanks.  Yea, ended up having to failover to the upgraded FTD and do the upgrade on the 6.2.3 unit via the CLI.  I'll keep an eye out for this the next time I need to do another upgrade on the pair.