Cluster Shared Volume in redirected access — everything slows down
A node has lost its direct path to the storage and is sending all its I/O over the cluster network via the coordinator node.
What you see
Virtual machines on one node become slow while the same machines run normally elsewhere. The volume shows redirected access in the cluster console.
What is actually wrong
A lost storage path on that node, a backup product holding a snapshot, or a filter driver that forces redirection.
Codes and articles
Start here — find out which fix applies
A script that runs the 1 inspection command from the write-up below and prints what each one returned. It reads the machine and changes nothing — every command that could write, delete, start or stop is excluded from it by construction. Run this first, then pick the fix its output points at.
No warranty. Use at your own risk - JbTecWiz accepts no liability. Read it before you run it, and have a backup.
Fixes (2)
Restore the node's direct path to the storage
One node is redirected.
Confirm which volumes and nodes are affected.
Get-ClusterSharedVolume | Format-Table Name,State,OwnerNodeGet-ClusterSharedVolumeState | Format-Table Name,Node,StateInfo,VolumeFriendlyName,RedirectedIOReason
RedirectedIOReason names the cause directly — NoDiskConnectivity, UserRequest, BlockRedirectedIOdueToFileSystemFilter and so on. That one field decides which of the fixes applies.
For NoDiskConnectivity, check the storage paths on that node.
Get-Disk | Format-Table Number,FriendlyName,OperationalStatus,BusTypempclaim -s -dGet-InitiatorPort | Format-Table NodeAddress,ConnectionType,OperationalStatus
Check the SAN zoning and the HBA or iSCSI sessions from that node specifically.
Turn redirection off once the path is restored.
Get-ClusterSharedVolume 'Cluster Disk 1' | Set-ClusterSharedVolumeState -State NormalIf it was set manually for maintenance and forgotten, that is the whole cause.
Get-ClusterSharedVolumeState | Where-Object StateInfo -ne 'Direct' | Format-List
Get-ClusterSharedVolumeState | Format-Table Name,Node,StateInfoNo warranty. Use at your own risk - JbTecWiz accepts no liability. Read it before you run it, and have a backup.
Find the filter driver forcing redirection
Every node is redirected.
List the filter drivers attached to the CSV volume.
fltmc filtersfltmc instances -v C:\ClusterStorage\Volume1
A backup agent or antivirus filter that is not CSV-aware forces the whole volume into redirected mode for as long as it is attached. The list names it in one command.
Check whether a backup is running or a snapshot has been left behind.
vssadmin list shadowsGet-ClusterSharedVolumeState | Format-Table Name,Node,StateInfo,RedirectedIOReason
Remove an orphaned snapshot if a backup failed part way.
vssadmin list shadows /for=C:\ClusterStorage\Volume1Add the CSV paths to the antivirus exclusions along with the Hyper-V process and file types — this is a documented requirement, not a performance tweak.
Get-MpPreference | Select-Object -ExpandProperty ExclusionPathUpdate the backup agent to a version that supports CSV direct I/O, or schedule its work in a maintenance window.
Get-ClusterSharedVolumeState | Format-Table Name,Node,StateInfoNo warranty. Use at your own risk - JbTecWiz accepts no liability. Read it before you run it, and have a backup.
Where this stops. This write-up was written and checked by hand. It says what each step changes, how to confirm it worked and how to reverse it, and anything destructive is flagged before you reach it. If it does not match what your machine is doing, search the Support Centre for the exact code or message — and when something needs a person, get in touch.