Fix 0X000013DD Cluster GUM Not Locker Error

Quick fix for cluster GUM lock errors: check cluster health, restart services or nodes, and verify network stability. This error usually appears during node restarts or network hiccups.

Quick Answer

Restart the Cluster service on the affected node(s) or reboot the node. If that doesn't clear the error, check for network instability and remove/re-add the node from the cluster.

What's Going On Here?

This error is a classic Windows Server Failover Cluster (WSFC) headache. The GUM (Global Update Manager) is the internal brain of your cluster that coordinates updates across all nodes. When a node that doesn't hold the GUM lock (the 'locker') tries to reserve it for a global update, you get 0X000013DD. Think of it as two people trying to grab the same pen at a meeting—one person is supposed to hand it over, but the other grabs it anyway.

In my six years running a help desk blog, I saw this most often after a node reboot or during a network blip. The node's cluster service gets confused about who holds the lock and tries to force a reservation. It's infuriating because the cluster might still look healthy, but you'll see this error in the cluster log and possibly in the event viewer.

Step-by-Step Fix

Check cluster health first. Run Get-ClusterNode | Select Name, State in PowerShell as admin. Look for nodes in 'Down' or 'Paused' state. If any are down, bring them up or evict them after checking hardware.
Restart the Cluster service on the affected node. On the node showing the error, open an admin PowerShell and run Restart-Service ClusSvc. This is the quickest fix—it resets the GUM state. If the node is the current lock holder, restarting it will fail over the cluster role briefly, so do this during a maintenance window.
Reboot the node if the service restart doesn't work. A full reboot clears any hung threads in the cluster stack. I've seen stubborn cases where only a reboot cleared the error.
Verify network configuration. Check that all nodes are on the same subnet and the cluster heartbeat network is responsive. Use Get-ClusterNetwork to list networks and confirm the heartbeat network (usually the one with 'Heartbeat' enabled) is healthy. If you see high latency or packet loss, fix your switches or NICs.
Evict and re-add the node. If the error persists, evict the problematic node: Remove-ClusterNode -Name NodeName (run from another node). Then repair the node, rejoin it with Add-ClusterNode -Name NodeName. This is a nuclear option but it forces a clean GUM state.

If That Doesn't Work

Sometimes the error is a symptom of a bigger issue. Here's what else to try:

Check for duplicate IPs. A rogue DHCP server or misconfigured static IP can cause intermittent cluster communication. Verify no two nodes have the same IP on the cluster network.
Review the cluster log. Run Get-ClusterLog -Destination C:\Logs and look for entries around the time of the error. You'll often see a chain of failures leading to the GUM issue.
Patch your Windows Server. I've seen this error fixed by cumulative updates on Server 2016 and 2019. Check for pending updates and install them.
Disable IPv6 if unused. Some cluster issues stem from IPv6 link-local confusion. If your cluster doesn't use IPv6, disable it on the cluster adapters.

Prevention Tips

You don't want this error to become a regular visitor. Here's how to keep your cluster stable:

Monitor network latency between nodes. Set up alerts for ping times exceeding 10ms on the heartbeat network.
Use quality NICs and switches. Cheap hardware causes intermittent failures that manifest as cluster errors. Your cluster deserves reliable gear.
Stagger node reboots. Never reboot all nodes at once—that's asking for GUM chaos. Always reboot one node, wait for it to fully join, then move to the next.
Keep your server updated. Regularly apply Windows updates to all nodes, preferably in a rolling fashion.

Pro tip from my blog days: Always document your cluster's expected heartbeat network. When things go sideways, you'll want to know which network is supposed to be handling cluster traffic.

This error won't take down a properly configured cluster, but it's a warning sign. Treat it as your cluster's way of saying 'check my connections'. Ignore it, and you'll eventually face a full cluster outage.