we're encountering an occasional issue in our client's environment where a task will be running several times longer than the job itself, making the task nonkillable and nonpausable. The job itself usually finishes in less than minute, while the task randomly runs for anywhere from two minutes to sometimes ten or even more minutes. The task is set to restart every minute.
When checking the logs we don't see anything significant that would explain the delay.
Can anybody provide any tips on how to analyse what's causing the delay and how to get the task to end as soon as the job ends?
