0

When installing a SQL patch on a 2-Node Automatic FCI with a shared disk (in the same datacenter / same subnet), with quorum achieved by Node and FileShare Majority, is it possible to install a SQL patch on one node without impacting the other node in that 2-Node Cluster? Obviously, this is for rollback purposes. What is the common approach in this scenario?

Thanks for the help!

0

The way we go about SQL patching and windows updates for a 2 node cluster is.

  1. In Failover Cluster Manager, remove the node you want to patch from the Possible owners (makes sure it cant failover to this node until you are ready)
  2. Patch the inactive node
  3. Re-add the node to the Possible owners
  4. Failover to the inactive node
  5. Leave this running for a day or two to make sure there are no issues
  6. Remove the first node from the Possible owners (now the inactive node)
  7. Patch the now inactive node
  8. Re-add the node to the Possible owners
  9. Fail-back

As long as your inactive node does not have the Quorum or Cluster Group patching or rebooing the inactive node should have zero impact.

The easiest way to make sure the node you want to patch does not own anything is using Powershell.

  1. Open Powershell
  2. Run the following command "import-module failoverclusters"
  3. Then run "get-clustergroup"

The output will show you what resources you have in the cluster and what node the currently live on.

Here is an article from MS on the topcic, https://support.microsoft.com/en-au/kb/958734.

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.