Hello all!
I have an Always On AG with 3 nodes: 2 synchronous and 1 asynchronous.
I have done a test where I stopped the 2 synchronous nodes and did a manual forced failover to the asynchronous one.
When I reconnected my 2 synchronous nodes and resumed the data movement on them with the AG, the state of the availability database on the synchronous nodes were stuck to SYNCHRONIZING state, even days after.
So it is impossible to fail back to the synchronous nodes.
I also checked, the synchronization state is healthy, and the last LSN commit is the same between all nodes. So I don’t understand why we are not in a SYNCHRONIZED state.
I have tried to remove the 2 synchronous nodes from the AG and re add them, or to remove the availability database from the AG and re add it, nothing is working.
The amount of transaction logs is low as well so I don’t understand what is wrong.