After issuing TERM signal to the Celery worker twice (Warm shutdown and cold shutdown) using Ctlr-C keyboard interrupt, the Celery worker is just hung up. It's not consuming messages or executing tasks (as expected) but it's not shutting down either. I ran strace on the Celery processes to see what's going on behind the scene. Here's the strace output on the PID of the Celery main process strace -p 27867 Process 27867 attached - interrupt to quit futex(0xb966a78, FUTEX_WAIT, 0, NULL And here's what I found doing strace on the child processes: strace -p 27874 Process 27874 attached - interrupt to quit select(4, [3], NULL, NULL, {0, 562000}) = 0 (Timeout) futex(0x871a808, FUTEX_WAKE, 1) = 0 select(4, [3], NULL, NULL, {1, 0}) = 0 (Timeout) futex(0x871a808, FUTEX_WAKE, 1) = 0 ...................................................... I know I could issue a KILL signal to the processes to get rid of them. But I'm curious to learn what is actually preventing these processes to be shutdown, and if it's possible to do something about it. Software stack: Python 2.6.2, Celery 2.4.6, CentOS 5.0 UPDATE: The CPU usage is down to almost 0%. These tasks are quite CPU intensive, so this confirms no tasks are currently active. Continue reading...
