Proper Way To Add Different Subnet Node to existing WSFC with FCI instances

  • Howdy,

    What I'm trying to do is setup an AO AG group between an FCI sql instance on one location (subnet) and a stand-alone SQL instance on a a different location (subnet). The different locations (subnets) have the own local storage.

    What is the proper way to add a node with a stand-alone SQL instance installed that is on a different subnet to an existing 2-node WSFC with an FCI set to use a disk witness?

    When I use the "Add Node" action on my existing WSFC instance and run the configuration validation, I get validation failures that I don't get when running the cluster validation on the existing instance. I get errors that the disk resources cannot connect to the new node. This is what I'm intending to do. The intention is that the existing FCI SQL instance will not failover to the new node. There will be an AO Availability Group that will span the two locations (subnets) and keep the different disks in synch.

    The acutal error messages in the validation check are ...

    DSM named Microsoft DSM, installed on node Node1Subnet1 is missing on node Node1Subnet2.

    DSM named Microsoft DSM, installed on node Node2Subnet1 is missing on node Node1Subnet2.

    Element not found

    It makes me think this is not the proper way to add the node.

    I've read through these articles and I'm ignorant enough not to see the answer. http://www.sqlservercentral.com/articles/FCI/92196/ is where I'm trying to get to. These MS Blogs have several detailed migration steps articles that all discuss what to do when setting up new clusters http://blogs.msdn.com/b/sqlalwayson/.

    ...thanks!

  • I've found this MSDN question with an answer to do the add node process without the storage check. Does this seem like the correct process to follow?

    http://social.msdn.microsoft.com/Forums/en-US/461a6bc1-f36b-448e-98b3-e72b503fb45f/3node-multisubnet-fciag-alwayson-cluster?forum=sqldisasterrecovery

  • The issue is most likely to be that the cluster wizard is automatically adding the new nodes local storage as cluster resources. Been a while but if I remember correctly there should be an option to add all eligible storage which you need to uncheck

    -----------------------------------------------------------------------------------------------------------

    "Ya can't make an omelette without breaking just a few eggs" 😉

  • I ended up proceeding with the add node process and choosing not to run the storage tests as part of the validation process. The node seems to have been added fine. There were some new disks added that could have come from the Site 2 node, although they are attached to one of the Site 1 nodes.

    Just fyi, I did not see an option to choose (or not to choose) to attempt to add that node's storage to the cluster. 😉

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

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