Availability Group - Core cluster resources switch to being owned by secondary node after windows patching reboots
Some of our AG clusters (async, manual failover) shift the ownership of their core cluster resources (e.g. cluster IP) from the primary node to the secondary node, which is in a different datacenter, after both nodes are restarted after monthly Windows patching. I can easily switch the resources back to the primary node in Cluster Manager, but is this something I should be worried about, like if the secondary datacenter goes offline? The quorum will still be intact in that case, but is that core cluster IP address relevant to the cluster and AG functioning? Thanks!
irl last edited by
[...] but is this something I should be worried about, like if the secondary datacenter goes offline?
That entirely depends on the design of the cluster. If there is a scenario where the core cluster resources reside in a location where a network outage between locations would lead to a potential split brain, then yes it would matter. Otherwise, it most likely would not.
The quorum will still be intact in that case, but is that core cluster IP address relevant to the cluster and AG functioning?
The ability to achieve quorum is unknown at this point. The cluster name, which acts as a common administrative endpoint for cluster operations, may be unavailable for a short period of time to regroup the cluster, assuming a whole lot. Really it entirely depends on the design and configuration of the cluster.
In terms of the AG functioning, it requires the cluster. No functioning cluster, no AG. Obviously, any nodes unable to communication will be a problem.