1

I am writing an ansible playbook that will go and kill PIDs.

I found a good answer that was relevant to me: https://stackoverflow.com/a/46541018/8863970

However, the first step which is Get running processes in my case I have two processes as follows:

# when i do: ps -ef | grep appp.py

ubuntu   17765     1  2 12:14 pts/0    00:00:04 python appp.py 
ubuntu   17784 17765  4 12:15 pts/0    00:00:05 /home/ubuntu/venvs/myvnv/bin/python /home/ubuntu/deploy/appp.py 
ubuntu   17844 14784  0 12:17 pts/0    00:00:00 grep --color=auto appp.py

It kills 17765 and then fails with:

failed: [10.10.1.1] (item=17784) => {"ansible_loop_var": "item", "changed": true, "cmd": "kill 17784", "delta": "0:00:00.002196", "end": "2020-04-26 12:23:22.833284", "item": "17784", "msg": "non-zero return code", "rc": 1, "start": "2020-04-26 12:23:22.831088", "stderr": "/bin/sh: 1: kill: No such process", "stderr_lines": ["/bin/sh: 1: kill: No such process"], "stdout": "", "stdout_lines": []}

Meaning... it can't find 17784 to kill.

Question is: How do I get only the parent PID (17784) with ps grep and then pass it on to kill it?

Saffik
  • 911
  • 4
  • 19
  • 45
  • 2
    Maybe just have your python app write a pid somewhere and use that, rather than trying to find it using `ps` and `grep`. Or write a systemd unit to run your python app so you can just use `systemctl stop ...` to stop it. – larsks Apr 26 '20 at 12:56
  • 1
    Guess what.... If you fix that first problem, it will error again on next pid 17844 corresponding to the current grep command you're running witch does not exist anymore. I suggest your go back to the question your linked in first place and read the second answer which solves your problem => use `pkill` rather than `grep + kill`. You can test your expression with `pgrep` first. Meanwhile the best solution is to follow @larsks comments above. – Zeitounator Apr 26 '20 at 13:00

1 Answers1

0

Tac this to the end | grep -v grep

Carl Andrews
  • 61
  • 1