Experienced unexpected behavior at customer site. Wondering if this is expected behavior or a potential bug. Had a continually running task with the default performance settings (that is consume all cores); and I understand this is not a good idea but is the unfortunate default. Also had scheduled task that is configured to use a couple cores.
What was found is that when the scheduled task kicked in, it pre-empted the continually running task and was able to run. The continually running task remained in the running state, but did not proceed even after scheduled task completed.
So first, what is the determining factor around what task gets what resources? And shouldn't any starved task resume once resources are available?