I have a few long-running python that run with many concurrent tasks in Deadline. When a failure occurs due to too many errors and all tasks are failed… I want to make sure I’m understanding the logic so that I can better clean up when failure occurs.
From what I read in past posts, Deadline uses a sigkill first to kill job when a job is failed either due to user triggered failure or due to hitting a job error limit for other tasks? Is that right?