Post reply

Added node to clustered sql prior to installing SQL on it, what to do :)

  • May 21, 2013 at 4:33 pm

    #278100

    Hi -

    I am suppose to setup an active/active clustered server (none of servers had any sql on it, these are new servers). The two servers are clustered at the windows level, I first setup node 1 as a clustered server following Microsoft documentation on how to setup active/active clustered server. When the SQL clustered server on node 1 was complete, I went ahead and ran the setup to add sql to clustered server. As I was going through the screens, I did not see where I could specify the sql network name and instance name for node 2. I went through screens, it showed node 2 was successfully added...But I only can see node1 under services and application in failover cluster manager.

    My question here is that 1) do I need to install SQL Server on node 2 prior to adding it to the clustered server, and if so, what option to select? and 2) do I need to remove node2 first(which showed node 2 machine name, and instance01 (that is what I specified as my instance name while I was setting up node 1)?

    Thanks.

  • May 21, 2013 at 10:51 pm

    #1617256

    It sounds like you've setup an active-passive cluster with a single instance of SQL Server running on node 1. Node 2 is part of the cluster but not running anything.

    You need 2 instances of SQL in an active-active cluster, one instance on each node.

    The sql installer has 2 options, one to create a new failover cluster and one to add a node to an existing failover cluster. You need to create 2 failover cluster instances and add a node to both.

  • May 22, 2013 at 6:42 am

    #1617393

    Do I need to 1) remove node2 which I did add after I set up node 1 cluster, 2) install a failover cluster on node 2, and then add node 1 to node 2 and node 2 to node 1 to make them active/active? Can you please tell me in sequence what the steps are?

    Thanks,

    Lava

  • May 22, 2013 at 5:19 pm

    #1617644

    Yes that is the correct order of steps, however you dont need step 1, it's fine the way it is. What you are doing is creating 2 failover cluster instances (FCI) on 2 physical nodes. It's the same as creating 1 FCI twice. You've already created one of the failover cluster instances, now you can create the other one. They are completely seperate but just happen to be running on the same hardware.

Viewing 4 posts - 1 through 3 (of 3 total)

You must be logged in to reply to this topic. Login to reply