1

I'm building up a SQL Server AlwaysOn AG and I'm on the step "failover", manual and automatic:

I noticed that in the Windows Server Failover Cluster a "Current Host Server" exists. My knowledge says that this is the "master" of the cluster that the cluster hosts. But now I noticed that the "Owner Node", which hosts the Availability Group(-role) a other node could be. (The owner node is also the primary instance of the AG, right?)

Now for me there exist two questions for the normal operating of an AlwaysOn AG:

Q1: Should the "Current Host Server" the same as the Owner Node of the AlwaysOn AG?

Q2: Imagine if I want to maintain a host. Shall I first failover the AG (via SSMS or with ALTER AVAILABILITY GROUP...) to the second node/another node and at second a failover of the WSFC-Current Host?

Maybe someone could help me with this questions and give me some advice.

Thanks

5

Should the "Current Host Server" the same as the Owner Node of the AlwaysOn AG?

That's the node that is currently hosting the core cluster resources, it isn't the same and is just another role. Different roles can be owned by different nodes. In this case, they do not need to be the same - and generally most people don't care who owns the core cluster resources in a WSFC that only has SQL Server. It does make a difference in certain situations but for the vast majority, it makes no real difference.

Imagine if I want to maintain a host. Shall I first failover the AG (via SSMS or with ALTER AVAILABILITY GROUP...) to the second node/another node and at second a failover of the WSFC-Current Host?

If you're going to be maintenance to a host, I'd move all roles (drain) off of that node. So, yes, everything should be moved from it.

  • So if Node1 is the primary instance, this Node1 receives all requests of the AG, right? But the "Current Host Server" doesn't mean that this host receives requests, or do I misunderstand?! – ssd_rider Dec 25 '17 at 16:26
  • The primary is the node that owns the AG resource. That's all that plays into the AG. Other roles have nothing to do with the AG. @ssd_rider – Sean Gallardy Dec 25 '17 at 22:38
  • So for example lets say there are 3 Nodes - Node1 and Node 2 Synchronous Mode and automatic failover where as Node 3 Different subnet asynchronous mode. The current host server is Node 3 where as the AG primary is Node1. Now what happens when Node 3 goes down will the Current host server fails to any of the other or the whole cluster goes down? – SQL_NoExpert Apr 12 at 19:18
  • @user9394033 1) You could easily test this. 2) It's role, so like any role it fails to another node. 3) We're missing a bunch of information so I'll go with all default settings, in which case it will stay up. – Sean Gallardy Apr 12 at 20:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.