Hi,
I have implemented AlwaysOn feature between two standalone SQL Server instances hosted on two Clustered nodes in two different subnets (multisite clustering with Node and File Share Majority quorum). I have configured AlwaysOn for automatic failover between the primary and the only secondary replica. There are two databases in it. The implementation did went successful.
Now, I before going live, I wanted to test the failover scenarios. First one was to manually failover between the nodes from SQL Server side, as well as from Failover Cluster Manager consoles. Both went perfect.
But, the second test was to stop the SQL Server service, and see the result. When I stopped the primary, the resource group in the Cluster Manager failed, and the databases were also not connecting. I expected the Availability group should have failed over to the secondary node, and the db's should have been up and running.
Did I miss something in the implementation part above, or it is an expected behavior of AlwaysON? If yes, then what does automatic failover imply?
Thanks & Regards