The timeout I added in #394 doesn't fully work: the containers survive the timeout, even if ocluster sends a message to ocurrent saying "Timeout (120 minutes)".
My intuition is that the worker successfully terminates, but doesn't (successfully or at all) kill the docker containers it spawned.
The timeout I added in #394 doesn't fully work: the containers survive the timeout, even if ocluster sends a message to ocurrent saying "Timeout (120 minutes)". My intuition is that the worker successfully terminates, but doesn't (successfully or at all) kill the docker containers it spawned.