What would be a proper order of operations when applying an OS patch on a SH cluster?

[danielbb]

In the last patch operation on our ITSI cluster of three SHs, we had syncing issues with the SHs. What can be a good sequence of operations when applying an OS patch on the cluster's servers?

[masonmorales]

Patch one server at a time. Starts with the members then do the captain last.

[danielbb]

Ok, when applying the patch, we simply shut down the SH and then start it after the patch is being applied?

[danielbb]

Looking at the upgrade procedure and it says at Upgrade a search head cluster

I would assume the sequence to be quite the same.

In the Cluster Administration it was said however -

-- The captain is the final member to restart (during a rolling-restart) and automatically invokes captaincy transfer, thus preventing captaincy from changing during the restart process.

That's exactly what @masonmorales just said.

[danielbb]

Should we apply a manual detention on each SH during the process?