2012R2 cluster csv - no access to storage and in redirected mode

Hi,

I have a two node 2012R2 cluster backed by a shared SAS JBOD. The cluster passes all validation tests - most specifically it passes the storage checks with flying colours.

It's using shared storage spaces which was configured and working perfectly with a quorum vdisk. The quorum can be moved between nodes without issue when moving the Storage Pool owner.

When adding a new vdisk to clustered shared volumes, immediately one of the nodes goes into redirected mode reporting that all storage paths are inaccessible. Snippet of the event; 

Cluster Shared Volume 'Volume1' ('Cluster Virtual Disk (Test)') is no longer directly accessible from this cluster node

If I change the CSV owner, the bad node stops redirecting and the other starts redirecting.

To visualise the physical config - it looks like this;
Image may be NSFW.
Clik here to view.

My MPIO paths are configured, and the devices show up correctly with all paths available (on both nodes)

Image may be NSFW.
Clik here to view.

Extraction from the Get-ClusterLog - I think I've found the part that shows exactly what's happening - but I don't know what it means...

00000a98.00000b10::2013/09/10-19:55:30.842 INFO  [RES] Physical Disk <Cluster Virtual Disk (VirtualisationStorage)>: HardDiskpQueryDiskFromStm: ClusterStmFindDisk returned error=2147943568
00000a98.00000b10::2013/09/10-19:55:30.842 ERR   [RES] Physical Disk <Cluster Virtual Disk (VirtualisationStorage)>: Failed to locate disk, status -2147023728
00000a98.00000b10::2013/09/10-19:55:30.842 ERR   [RES] Physical Disk <Cluster Virtual Disk (VirtualisationStorage)>: FindTheDiskNumber: Unable to find disk. Error: '2147943568'.
000005ec.00000908::2013/09/10-19:55:30.842 INFO  [RCM] Group da40e60c-7971-4323-bd45-3a19f36599db: done going through resources, returning true
000005ec.00000908::2013/09/10-19:55:30.842 INFO  [RCM] da40e60c-7971-4323-bd45-3a19f36599db should not fail back from 2 due to possible owners
000005ec.00000908::2013/09/10-19:55:30.842 INFO  [RCM] da40e60c-7971-4323-bd45-3a19f36599db should not fail back from 2 due to paused nodes
00000a98.00000b10::2013/09/10-19:55:30.842 ERR   [RHS] Error 2147943568 from ResourceControl for resource Cluster Virtual Disk (VirtualisationStorage).
000005ec.000005b4::2013/09/10-19:55:30.842 ERR   [DCM] failed to get diskNumber, status 1

Thanks and regards,

Matt.