Minimal code to reproduce a weird issue with leftover defunct processes from python's multiprocess.pool() in spawn mode.

The code generating the defunct processes is in multi.py and is the actual pool() example just with added code to enable spawn mode.

To actually see the zombies the whole thing is run from another python program.

Things work correctly in a normal environment or when the running process is aware of SIGCHLD and handles it by wait(2)ing.

Build

docker build -t multi .

Run without issues form bash

$ docker run -it multi bash main.sh
[1, 4, 9]
UID PID PPID C STIME TTY TIME CMD
root 1 0 3 11:31 pts/0 00:00:00 bash main.sh
root 10 1 4 11:31 pts/0 00:00:00 /usr/local/bin/python -c from multiprocessing.resource_tracker import main;main(5)
root 19 1 0 11:32 pts/0 00:00:00 ps -ef --forest
[1, 4, 9]
UID PID PPID C STIME TTY TIME CMD
root 1 0 1 11:31 pts/0 00:00:00 bash main.sh
root 22 1 0 11:32 pts/0 00:00:00 /usr/local/bin/python -c from multiprocessing.resource_tracker import main;main(5)
root 31 1 0 11:32 pts/0 00:00:00 ps -ef --forest
[1, 4, 9]
UID PID PPID C STIME TTY TIME CMD
root 1 0 1 11:31 pts/0 00:00:00 bash main.sh
root 34 1 0 11:32 pts/0 00:00:00 /usr/local/bin/python -c from multiprocessing.resource_tracker import main;main(5)
root 43 1 0 11:32 pts/0 00:00:00 ps -ef --forest
...

Running from python produces defunct processes

$ docker run -it multi python main.py
[1, 4, 9]
UID PID PPID C STIME TTY TIME CMD
root 1 0 11 11:33 pts/0 00:00:00 python main.py
root 8 1 0 11:33 pts/0 00:00:00 [python] <defunct>
root 17 1 0 11:33 pts/0 00:00:00 ps -ef --forest
[1, 4, 9]
UID PID PPID C STIME TTY TIME CMD
root 1 0 6 11:33 pts/0 00:00:00 python main.py
root 8 1 3 11:33 pts/0 00:00:00 [python] <defunct>
root 19 1 0 11:33 pts/0 00:00:00 [python] <defunct>
root 28 1 0 11:33 pts/0 00:00:00 ps -ef --forest
[1, 4, 9]
UID PID PPID C STIME TTY TIME CMD
root 1 0 4 11:33 pts/0 00:00:00 python main.py
root 8 1 1 11:33 pts/0 00:00:00 [python] <defunct>
root 19 1 3 11:33 pts/0 00:00:00 [python] <defunct>
root 30 1 0 11:33 pts/0 00:00:00 [python] <defunct>
root 39 1 0 11:33 pts/0 00:00:00 ps -ef --forest
[1, 4, 9]
UID PID PPID C STIME TTY TIME CMD
root 1 0 3 11:33 pts/0 00:00:00 python main.py
root 8 1 1 11:33 pts/0 00:00:00 [python] <defunct>
root 19 1 1 11:33 pts/0 00:00:00 [python] <defunct>
root 30 1 4 11:33 pts/0 00:00:00 [python] <defunct>
root 41 1 0 11:33 pts/0 00:00:00 [python] <defunct>
root 50 1 0 11:33 pts/0 00:00:00 ps -ef --forest
...

Running from SIGCHLD-aware python program avoids the issue

$ docker run -it multi python main-sig.py
[1, 4, 9]
UID PID PPID C STIME TTY TIME CMD
root 1 0 11 11:34 pts/0 00:00:00 python main-sig.py
root 18 1 0 11:34 pts/0 00:00:00 ps -ef --forest
[1, 4, 9]
UID PID PPID C STIME TTY TIME CMD
root 1 0 5 11:34 pts/0 00:00:00 python main-sig.py
root 29 1 0 11:34 pts/0 00:00:00 ps -ef --forest
[1, 4, 9]
UID PID PPID C STIME TTY TIME CMD
root 1 0 3 11:34 pts/0 00:00:00 python main-sig.py
root 40 1 0 11:34 pts/0 00:00:00 ps -ef --forest
[1, 4, 9]
UID PID PPID C STIME TTY TIME CMD
root 1 0 3 11:34 pts/0 00:00:00 python main-sig.py
root 51 1 0 11:34 pts/0 00:00:00 ps -ef --forest
...