An administrator deleted a slave VM which the vSphere plugin had created dynamically.

After that happened, the vSphere plugin continues to believe that this slave is in its list, and is not removing it.

Our total number of slave instances is capped – when this happened, we were unable to continue until the cap was raised, as we continued to get this in the logs after the even occurred:
There are 1 number of VMs in this bucket. The instance cap for the bucket is: 1

Restarting Jenkins did not resolve the problem. The workaround was to increment our cap to 2, so that one slot was still available.

Is there a way to clear the current list of slaves the plugin believes exists?