This is somewhat anecdotal, but twice now we've seen an issue where if a job is queued for one of our macOS nodes, and the node is restarted, the node will forever stick at "node is starting up". To fix this, you actually have to restart the jenkins server (or service) as it never figures out the node has started. Seems like there's a race condition here?

Node is on "Unix agent, version 3283.v92c105e0f819"

Server is on 2.492.1

I'm probably leaving out info. We haven't been able to find any actual logging that indicates what's going on, but the fix always seems to a required restart of Jenkins (not the node). 

I did not see another report of this, apologies if I just didn't search well enough. 

Thank you.