Hi,
My previous experience with failover clustering has only been with hyper-v clustering, so I am looking for some advice on how SQL clustering should be configured and how it works.
with hyper-v clustering i can move a "role" (a virtual machine) using live migration with no downtime
if a physical node goes down, the VM that runs on top goes down and fails over to another node
part of our quorum setting involved having either a disk quorum or a file share witness
now onto SQL, which i know very little about:
we have multiple instances running on a node (roles), if we perform a planned failover all the databases go offline and then come online on the other node, a planned failover incurs some downtime - is this normal?
if a physical node goes down, DBs go offline and come back online on the other node - i assume this is normal
we dont seem to have a disk or file share witness but we do have a DTC role - is this normal for quorum settings? it seems it is in node majority for the quorum, but we have had failures (in a 2 node cluster) and we can't connect to the cluster manager until the node that went down comes back on, yes this could be due to networking but i'm trying to rule out and understand the quorum setting first. if we perform a planned failover and pause the node (draining the roles) then take the node down, we can successfully still manage the cluster running on the remaining node.
The failover cluster nodes are running Windows Server 2012 R2 Standard, and SQL is 2012 R2 Standard, appreciate any advise you can provide.
Many thanks for your time
Steve