I guess this is probably supported, but not documented...

I want to configure a pool of single-use dynamic workers:
once a jenkins tied to that pool/label is about to run, a new node is created in Openstack cloud, the job runs, then the node is destroyed.

I see "retention time" of -1 means "keep nodes forever". I'm guessing setting it to 0 would do something similar to what I described. but not sure... because if nodes are cleaned up once a minute (for example), then another job can start running on that node in those 60 seconds between the last job finished and the cleanup starts..

so how do I force deleting nodes after each jenkins job?