Clustering helps you achieve higher availability by wiring up more than one server to the same set of shared drives. If one server has a catastrophic problem like failures in Windows, the motherboard, the memory, the network cards, etc, then SQL Server will automatically restart on another server connected to the same storage.
Clustering is a wonderful thing, but it changes the way you tweak certain settings in SQL Server like min memory. We have to make sure that when instances fail over, we can still run all of them on the same node.
We check sys.dm_os_cluster_nodes looking for entries. We don’t do anything fancy like checking the health of the cluster or which nodes are involved – we just want you to know you’re working on one.
We also alert you if no failover cluster nodes are available and ready to perform the failover.
To fix the problem, it’s time to sketch out your failover cluster, understand where the instance can fail over, and do regular failover testing to make sure the cluster can fail over when something goes wrong.