Hello team,
We are using attached white paper document for cross cluster always on availability group migration (HERE) and testing the suggested set of steps in our environment which is four node (Lets say Nodes 1, 2, 3, 4) windows 2008 cluster environment and as per requirement we have created another new destination windows 2012 cluster which is also having four nodes (Nodes 5, 6, 7, 8) .
Configuration of Source and Destination
WSFC cluster | Windows | SQL Server |
Source cluster | Windows Server 2008 R2 Enterprise. | SQL Server 2012 RTM (Enterprise) |
Destination cluster | Windows Server 2012 R2 Datacenter. | SQL Server 2012 SP1 (Enterprise) |
As per document we have executed below steps
- Creatednew testAG on primary replica (Node 1) in windows 2008 cluster environment and added test database to that AG with 1 secondary replica (Node 2 with synchronous commit). So there are two replicas in TestAG now (Node 1 & 2)
- Changed the HADR ClusterContext to windows 2008 cluster for Node 5 & 6 of destination cluster (Windows 2012 cluster).
- Added node 5 & 6 as secondary replicas of TestAG. Node 5 has been configured as synchronous commit replica whereas an asynchronous commit for node 6
- Then deleted the listener and set the TestAG to OFFLINE
- We tried to create new AG with same name on destination cluster node 5 but we were unable to add the test database (already synchronized ) to this new TestAG as that database is in restoring mode (so not accessible to add in AG)
As per my understanding we have followed all the steps in same fashion in which it has been described in the white paper. We require your kind support in removing this hurdle or you can advise any alternate method by which we can migrate our database between these two hardware set with minimum downtime.
Thanks a lot