0

I am exploring the 2 node cluster with file share.

The example is as shown here: https://docs.microsoft.com/en-us/azure-stack/hci/concepts/quorum#two-nodes-with-a-witness

enter image description here

Majority is 2/3. So when either node goes down, the cluster will stay up.

My question is - suppose the node 2 is restarted by an engineer as part of maintenance/patching without taking it out of the cluster, and the file share is unstable, then will the cluster dynamically adjust to survive with 1 node (primary) vote? Or will cluster go down?

I understand there is dynamic witness and dynamic node voting that should handle this? haven't found any documentation about this example - how dynamic vote kicks in when only node 1 and file share witness (2 of the originally configured 3) are alive.

3
  • Define "fileshare is unstable" ?
    – Shanky
    May 9 at 11:35
  • Why not try it? Set up three VMs in a cluster. Pause whatever nodes you wish and look what's being recored to the cluster log. The log explains with details what's really going on.
    – vonPryz
    May 9 at 11:35
  • File share unstable due to network issues I can see cluster log indicating inaccessible.
    – variable
    May 9 at 12:44

1 Answer 1

0

If the witness is not accessible, the cluster should readjust the vote so that only one of the 2 node gets a vote.

Knowing that, if you are lucky, the node that will be rebooted will not have a vote and the other node will remain up. If you are not and the rebooted node was the one with the vote, then the cluster will go down.

The big question is why is your witness unstable ? This is an important part of the cluster and it should not be "unstable" so I would put some time to fix this if I were you. After all, you build a cluster to have high availability and if you can't trust the witness, you can't say you have HA.

7
  • 1
    The vote should move on a controlled reboot. May 9 at 12:32
  • Vote moves to node based on luck?
    – variable
    May 9 at 12:40
  • @DavidBrowne-Microsoft - can you elaborate on your comment please?
    – variable
    May 9 at 13:02
  • Dynamic vote will remove the vote to one of the 2 nodes (I don't know exactly how it choose). If you reboot the node that has the vote, it will go down. If you reboot the note that doesn't have the vote, it will stay up. That's what I mean by " lucky" May 9 at 13:39
  • 1
    You could have 2 AG using the same cluster with AG1 being primary on Server A and AG2 being primary on Server B. It can't look at all the possibilities and try to update the vote for every specific condition... After all, the cluster is expecting your witness to be working fine. You should fix the issue (witness) instead of trying to go around all the problem that it could induce. May 9 at 19:17

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.