Fix ERROR_QUORUM_OWNER_ALIVE (0X000013AA) in Failover Cluster

Cause 1: Cluster service stuck in a holding pattern

The most common reason you see ERROR_QUORUM_OWNER_ALIVE (0X000013AA) is that the cluster service on the current quorum owner node won't release its claim. This happens after a network blip or a sudden reboot. The cluster thinks the owner is alive but it's actually in a ghost state. I've seen this on Windows Server 2019 and 2022 clusters with a file share witness.

Here's the fix that works 90% of the time:

1. Log on to the node that currently owns the quorum. You can find this in Failover Cluster Manager under "Nodes" — look for the one with the checkmark next to "Quorum". If you can't tell, run this PowerShell on any node: Get-ClusterQuorum. It shows you the current owner.
2. Stop the cluster service. Open an elevated command prompt (run as Administrator). Type: net stop clussvc. Wait for it to say "The Cluster service was stopped successfully."
3. Wait 30 seconds. Let the other nodes see the owner is gone. During this time, the cluster will try to elect a new owner. If you restart too fast, the original node might grab it again.
4. Restart the cluster service. Type: net start clussvc. After this, the node should rejoin the cluster cleanly.
5. Check the cluster health. Open Failover Cluster Manager, click on your cluster name. Under "Summary", you should see all nodes with a green checkmark. No more 0X000013AA errors in the event log.

What to expect after this step: If the error goes away, you're done. If it comes back after a few minutes, move to Cause 2 below.

Cause 2: File share witness points to a dead or unreachable path

If your cluster uses a file share witness and that share is on a drive that's disconnected or slow, the quorum owner can't confirm it's alive but the cluster still holds the lock. I've seen this with DFS shares or shares on a domain controller that's offline.

Here's the fix:

1. Find the witness path. Run on any node: Get-ClusterQuorum. Look for "FileShareWitness" and the path. It'll look like \\server\share.
2. Test the path. Open a File Explorer on the node that shows the error. Paste the witness path into the address bar. If you get "Access denied" or "Network path not found", that's your problem. If you can access it, skip to step 4.
3. Fix the witness location. Move the witness to a stable location. The safest bet is a dedicated file server that's always on, or a cloud-based witness. To change it, run: Set-ClusterQuorum -FileShareWitness \\new-server\witness-share. Replace the path with your new share. If you don't have a new share, you can temporarily use a disk witness if your cluster has a shared disk.
4. Force a quorum refresh. After changing the witness, run: Reset-ClusterVMMetadata (this clears stuck state). Then restart the cluster service on all nodes: Restart-Service clussvc on each one.
5. Test again. Wait 2 minutes. Check the event log for the error — it should be gone.

Cause 3: Network connectivity issues between nodes

Sometimes the quorum owner is actually alive but the other nodes can't reach it because of a firewall rule or a misconfigured VLAN. The cluster sees the owner but can't communicate, so it throws this error. I've fixed this more times than I can count on clusters with two network adapters (one for cluster traffic, one for management).

Here's the fix:

1. Ping the quorum owner's cluster IP. On a non-owner node, open Command Prompt. Type: ping <quorum-owner-IP>. If you get timeouts or "Destination host unreachable", the network is broken.
2. Check firewall rules. The cluster uses ports 3343 (TCP/UDP) and 445 (TCP). On the quorum owner, run: netsh advfirewall firewall show rule name="Cluster". If no rule exists, add one: netsh advfirewall firewall add rule name="Cluster" dir=in action=allow protocol=any remoteip=any localport=3343,445. Do this on all nodes.
3. Check network adapter bindings. On each node, open Network Connections (ncpa.cpl). Right-click the adapter used for cluster traffic, click Properties. Uncheck everything except "Internet Protocol Version 4 (TCP/IPv4)" and "Cluster Network Driver". This prevents conflicts from other protocols.
4. Test again after 30 seconds. Run Get-ClusterNetwork to see if the cluster shows all networks as up ("Up" status). If any show "Down", you've got a physical cable or switch problem.

Quick-reference summary table

Bottom line: Start with restarting the cluster service. It's quick, safe, and fixes most cases. If the error comes back, check the witness path. If that's fine, look at network connectivity. Don't try to rebuild the whole cluster — that's overkill for this error.