I have a 2 node cluster.
2 - server running clustering with file server service installed. both are inside vSphere 6.0u2 hosted in a UCS blade system. Both have 2 nics, 1 for client communications with a default route and 1 connected to iSCSI vlan with ip addy only and no default rout and iscsi is no network communications.
I have a top level cluster with a file share witness quorum, this is a network share connected to a different physical server in my domain, both cluster servers have read/write access to this share along with the top level cluster and the cluster role server.
When I pause the active node the cluster master and cluster role server is only ping'able from the network it is on, ie. 192.168.1.x all other vlans, that were able to hit it, before pausing the active node, are now disconnected form the shares and can not ping the cluster role server. the same happens if I pause the cluster node that the top level cluster is connected to.
I have destroyed this cluster a few time and have followed best practice while setting this up.
I don't believe this is switch related as we have no other problems connecting inter vlans in any of our other servers. I have servers in 5 subnets on that many vlans and no issues. only problem is when I pause one node it makes the cluster not reachable even though the cluster is online and reachable from its own subnet.
Getting ready to trash the whole concept and go back to one file server.
Any insight would be appreciated.
JR