Node Disk Resource Behavior

sstark85

What is the expected behavior of disks, volumes and mount point resources associated to a node when they no longer exist? Should re-discovery occur and remove them from the node or will they persist as stale resources on the node? A common example is a set of servers in a cluster that have a clustered shared volume, the mount will only exist on one node, say node A, when the cluster fails over the mount point will be moved to node B. I have noticed that this is not reflected properly in SolarWinds, the mount point will remain on node A, and sometimes it will never be discovered on Node B until a manual refresh of resources is performed.

mesverrum

When it comes to adding or removing monitored objects in Solarwinds it is safe to assume nothing is automatic.  You will need to delete the mounts from the node where they show down and you will need to list resources to add the drives on the newly active server.

As a general rule with clustered systems like that i would set up a node in solarwinds pointed at the shared cluster IP/dnsname, and only monitor my mounts on that node, not on the individual servers since a failover event would cause them to display incorrectly.  Anything that would be specific to one host or another such as network interfaces and local disks do not get added to the VIP node, and are only monitored on the individual nodes.