0

I found Questions regarding removing a secondary node from an availability group with Tony Hinkle's answer, but I'd like to get some clarification on 1) removing the databases from the AG, 2) removing the secondary node from the cluster, and 3) removing all quorum votes except for the one node in the cluster.

Does step 2 mean removing the node from the AG first, then from WSFC?

How does the file share witness factor into removing all the other quorum votes? Should the one remaining cluster node and the file share witness keep their votes or just the node?

Thanks!

Improve this question

1 Answer 1

Reset to default
0

removing the databases from the AG

Database can be removed from AG but if you want it back you cannot simply add it, you have to first take full and log backup from primary, restore it on the DB and then add it. For VLDB is this a challenge.

removing the secondary node from the cluster

Again it would be advisable to break the AG first and then remove the node. It should then be removed from WSFC, If you want to add it back restore all the databases and then add it back with same steps, first in WSFC then in AG.

removing all quorum votes except for the one node in the cluster.

If node is removed its vote will be removed. With one node what is point in worrying about Quorum.

How does the file share witness factor into removing all the other quorum votes?

With single node, like I said above what is point in FS witness anyway.

I would say I dont like this idea very much. This will create more work for you. Please note if you have SA purchased and you are NOT running any queries on secondary which means it non readable you do not have to license it, provided it has same no of cores as primary.

Improve this answer
2
  • Thanks for the reply. The removal of the secondary node is due to a datacenter move, not due to saving money on licensing...the node will be added back once the datacenters are ready. I'm worried about the AG and/or cluster and/or quorum thinking the old secondary is still in play, which could cause the entire AG to go offline , so I want to make sure the environment is configured appropriately. I like the idea of the cluster and AG sticking around even with just the primary node, as Tony suggested, then when ready I can simply add the new vm to the cluster and AG and seed and add the dbs.
    – Bobogator
    Commented Mar 31, 2022 at 13:16
  • If you are migrating to DC do you know you can add node to WSFC from other DC, or create separate WSFC and AG in new DC and configure AG between the WSFC's in 2 DC( DAG). Assuming you have stretched VLAN or some connectivity. Removing secondary and disturbing current AG would not be a good idea.
    – Shanky
    Commented Mar 31, 2022 at 13:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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