Forum Discussion

Michal_Kratoch1's avatar
Michal_Kratoch1
Icon for Nimbostratus rankNimbostratus
Feb 19, 2017

Standby shows as offline in HA pair in version 12.1.2

Hello all, I know this topic has been discussed before but for older code versions and none of the solutions worked for me and I tried pretty much all of them. I have 2 3900 series boxes in HA pair. Network failover, port lockdown set to "allow all", boxes directly connected in HA vlan, no device in between. Upgraded from 12.0.0 to 12.1.2 and in the "Overview" page on the active box it shows the standby peer as "offline". If I failover to the other box it will again show the standby peer as "offline" from the new active box. Everything looks fine from the standby box "Overview". Both failover group and trust group show full sync, HA status on both boxes shows "in sync", no issues there.

 

I tried multiple failovers, reboots from CLI, power cycles, changing port lockdown, even tried the process of forcing the configuration reload to no avail, still the same.

 

Does anybody experience the same behaviour? I suspect this is a bug in the 12.1.2 code version.

 

 

5 Replies

  • Hello Michal,

     

    I would have tried resetting the HA configuration. i.e, Break the HA and reconfigure it again. Make sure that you consider this option carefully if the devices are in production..

     

    Regards,

     

    Mohammed Shiraz

     

  • Hi all, I think we are hitting same issue in 12.1.3.3. Both Failover Network and Mirroring are populated, but the active device sees the standby as offline. Standby units sees the active one as active (green status).

     

    • ivanbermejocha1's avatar
      ivanbermejocha1
      Icon for Nimbostratus rankNimbostratus

      I answer on my own post, just if it helps. We did realize that HA interface was in "Allow None" in one of the devices. The curious thing is that the HA interface was "locked" in the active one, and the active one couldn't see the standby one, but the standby one could see the active one.

       

  • I had this exact situation happen in 12.1.3 on a Viprion pair. Doing a bigstart restart sod on the standby guest didn't have any effect, but doing it on the active guest (the one reporting the standby unit as offline) fixed the issue but I did experience a bit of strange behavior:

     

    I'm using vCMP and I have two guests with LTM running on both. Both had the same situation. When I restarted the sod on guest A, a failover happened and caused the standby guest A to go active. When I restarted the sod on guest B, there was no failover. Any ideas of why guest A failed over?