Cluster failover due to network failure | SQL Server Performance Forums
SQL Server Performance Forum – Threads Archive
Cluster failover due to network failureHi,
We have some network related issues and it caused the sql instances to fail over to passive node.We have 3 node a/p/a cluster setup.I moved the sql instance group from passive node to active node from Cluster administrator. Here I have couple of questions:
1.Is the SQL Server insatnce restart every time we move the groups from one node to other? Because I’m seeing in the error log that the sql server instance has restarted. Is this normal behavior?
2.We have active/passive/active cluster setup. Node1 should failover to node2 and node 3 should failover to node2 but NOT node1 to node3 and node3 to node1.We have lost network communication on node3 and it moved to node2 and node1 also moved to nod2. Later node1 instance moved back to node1 from node2. But node3 instance is still on node2. It did not move to its preferred node node3?please advice…
3. We have a backup group called Backup on node3. When the netwok communication is lost, sql group is moved node2 But the Backup group did not and all the backups of the sql instance are failing with the below error:BackupDiskFile::OpenMedia: Backup device ‘Z:\FULLBACKUPProd_06-24-2009.bak’ failed to open. Operating system error 21(The device is not ready.)what should I do to fail over this Backup group(contails only the Backup drive Z) along with sql group when ever failover is initiated inorder to avoid the Backup failures
4. Node 1 drives are visible from node3. Is this normal? before failover, node1 drives will not be visible from node3. Is this normal?or what should I do not to see node1 drives from node2?
When you configure the failover of the resource group, you can select some behavior like "failback inmediately the prefered node is available" , "Failback at xx hour" or "Don’t failback automatically"
Check what option do you have for this group.