Have setup some Av Groups, in a 2 node cluster + file share witness with 2 single instances on their own storage e.g. not shared clustered storage, and they fail over nicely etc if a server reboots or what ever, however, if the storage is removed(e.g. disconnect iscsi which is on a different volume to the operating system (so the OS keeps running but the SQL data volume disappears).. the server wont fail over , and the cluster breaks/stops , and the AV group stops too, so unable to connect to any instances via the listener.
Is this by design or is the cluster meant to fail over should the storage break?
Thanks