I have 3 servers on VMware workstation:

  1. Win 2012 R2 as ADDC -Cluster service.
  2. Win 2012 R2 as SQL server primary replica .
  3. Win 2012 R2 as SQL server secondary replica (auto fail-over).

    My Availability group fail in this scenario:

    When I disconnect primary replica from the network the cluster fail to announce a fail-over to secondary replica when I try to manual fail-over from secondary replica instance I get this error:

enter image description here however this not the case when I stop primary replica from cluster manager because it will do auto fail-over.

  • 1
    How many nodes you have ADDC-Cluster Service? You had configured between primary replica & secondary replica with " Synchronous-commit replicas" or "Asynchronous-commit replicas" . Aug 16, 2015 at 8:17
  • You have to check out before doing the manual failover.To support a manual failover, the current primary replica must be set to synchronous-commit mode and a secondary replica must be: (1)Configured for synchronous-commit mode. (2)Currently synchronized with the primary replica. otherwise you shall get error . for ref msdn.microsoft.com/en-us/library/hh213151.aspx Aug 16, 2015 at 8:41
  • I have 2 nodes in cluster. Synchronous-commit replicas.
    – danarj
    Aug 16, 2015 at 9:28
  • I can click on failover on the cluster node and it is working. and manual also not working in the above scenario.
    – danarj
    Aug 16, 2015 at 9:29

1 Answer 1


Ok I found what I was missed and it was the cluster because there is only two nodes in the cluster so I add third node and it is working now.

  • @danaraj, that's why i have asked first . how many you have nodes in your ADDC-cluster. Because before doing manual failover or force manual failover. you must have all quorum cluster information. Aug 17, 2015 at 13:26
  • @danaraj, In your case you had disconnect your primary replica and secondary replica is not working. so, in that case either you make (secondary replica as primary with manual failover) or you force manual failover to other nodes as primary replica , in this case some certainty of data loss. Aug 17, 2015 at 13:36
  • this solve my problem msdn.microsoft.com/en-us/library/… Microsoft recommend Node majority quorum in case you have odd number of nodes but in case you have even number you should use node and disk majority or node and file share majority
    – danarj
    Aug 17, 2015 at 14:42
  • @danraj,yes of course. That is one way. For sync of data. But you have conscious before adding or deleting the cluster node. And before that always take primary server node. Aug 17, 2015 at 14:45
  • but I after I disconnect primary replica the failover succeed but after recovery from the failover the other secondary nodes stop syncing
    – danarj
    Aug 17, 2015 at 14:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.