Failing over to DR site Availability group



  • I have a 4 node cluster SQL 2012 availability group database with one witness server at the primary site. 3 nodes are in the primary site and one node in the DR site.

    Due to maintenance at the Primary Site I need to failover to the DR site and making it as the primary.

    Here is my question, if I Failover to the Dr Site and turn off all the nodes one by one gracefully including the witness server, will the Dynamin Quorum take over and keep the cluster and the availability group alive or the Windows Server Failover Cluster goes down and AG is inaccessible and as a result I will have to bring the cluster service up with force quorum



  • I have a 4 node cluster SQL 2012 availability group database with one witness server at the primary site. 3 nodes are in the primary site and one node in the DR site.

    Pedantically, if you have a 4 node cluster with 3 nodes and 1 node, then you don't have a witness server. I don't know what your witness is, if any.

    Here is my question, if I Failover to the Dr Site and turn off all the nodes one by one gracefully including the witness server, will the Dynamin Quorum take over and keep the cluster and the availability group alive or the Windows Server Failover Cluster goes down and AG is inaccessible and as a result I will have to bring the cluster service up with force quorum

    While it's possible to make this work, it requires a setup that you're not currently running since there will be a communications failure between nodes. Should there be a proper witness setup that has communications even when the primary site is down, you can move the core cluster resources to the node in your DR and then shut down the other nodes one by one and achieve this. Since I doubt this is your architecture, most likely it won't work that way.




Suggested Topics

  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2