Fix 0X00001708 Cluster Node Evict Without Cleanup

Quick answer for pros

Run Remove-ClusterNode -Name NodeName -Force from an active cluster node, then clear the cluster computer object from AD if it's orphaned.

Why this happens

You kicked a node out of a Windows Server Failover Cluster (2012 R2, 2016, 2019, 2022 — all do this). The eviction itself succeeded — the node is gone from the cluster quorum. But the cleanup step failed. Cleanup means removing the node's virtual computer object from Active Directory, clearing its registry entries, and stopping the cluster service on that node. The culprit here is almost always one of three things:

The evicted node lost network connectivity before cleanup could complete
The cluster service account (the one running Cluster Service) doesn't have permission to delete the computer object in AD
The node's cluster database is corrupted and the cleanup script couldn't finish

I've seen this mostly when someone forces an eviction through Failover Cluster Manager while the node is half-dead. The UI does the evict, then tries to cleanup, and fails silently leaving you with a ghost node.

Fix steps

Open PowerShell as Admin on an active cluster node. Doesn't matter which one — pick the one that's healthy and has network access.

Run this command to see the bad node:

Get-ClusterNode | Where-Object {$_.State -eq 'Down'}

Force remove it:
```
Remove-ClusterNode -Name NodeName -Force
```
Replace NodeName with the actual node name. The -Force flag skips the cleanup attempt and just kills the cluster membership.
Check if the computer object is still in AD:
```
Get-ADComputer -Identity NodeName -ErrorAction SilentlyContinue
```
If it returns something, that object is orphaned. You need to delete it manually.
Delete the orphaned computer object:
```
Remove-ADComputer -Identity NodeName -Confirm:$false
```
This requires Domain Admin rights or delegated permissions to delete computer objects.
Clean up the evicted node itself: Log into that node (if you can), open Services.msc, find the Cluster Service, stop it and set it to Disabled. Then delete this registry key:
```
HKLM\Cluster
```
That wipes the local cluster database so it won't try to rejoin on reboot.

Alternative fixes if the main one fails

If Remove-ClusterNode gives you an access denied error: The cluster service account probably doesn't have enough rights. You can either grant it the 'Create Computer Objects' permission in the Computers OU, or use a different approach — run the removal from the evicted node itself (if it's still alive). Open PowerShell there and run:

Clear-ClusterNode -Cleanup

If the node is completely dead (no network, hardware failure): You can't cleanup the AD object from inside the cluster. Use AD Users and Computers directly, find the computer object under the Computers container or the cluster's OU, right-click and delete it. Then on a surviving cluster node, run:

Get-ClusterNode | Remove-ClusterNode -Name DeadNode -Force

Yes, you're removing a node that's already gone. This cleans up the cluster database locally.

If you still see the node in Failover Cluster Manager: Restart the Cluster Service on the active node you ran the command from. That flushes the cache.

Prevention tip

Don't evict nodes from the GUI when they're unreachable or dying. Always use Suspend-ClusterNode -Drain first, then shutdown the node cleanly, then evict it from a healthy node. And make sure your cluster service account has 'Delete Computer Objects' permission on the OU where the cluster computer objects live. Set it once in AD delegation and you'll never see this error again.