Kill supervisor on Docker when any of the services fail

README.md

Killing supervisor if any of it's child processes fail

The trick is to only register the listener for events that indicate failure, namely

• PROCESS_STATE_STOPPED
• PROCESS_STATE_EXITED
• PROCESS_STATE_FATAL

Once they do, we should send a SIGQUIT to Supervisor.

stop-supervisor.sh

#!/bin/bash

printf "READY\n";

while read line; do
  echo "Processing Event: $line" >&2;
  kill -3 $(cat "/var/run/supervisord.pid")
done < /dev/stdin

supervisord.conf

[supervisord]
nodaemon=true
loglevel=debug
logfile=/var/log/supervisor/supervisord.log
pidfile=/var/run/supervisord.pid
childlogdir=/var/log/supervisor

[program:nginx]
command=nginx -g "daemon off;"
redirect_stderr=true

[program:php-fpm]
command=php-fpm
redirect_stderr=true

[eventlistener:processes]
command=stop-supervisor.sh
events=PROCESS_STATE_STOPPED, PROCESS_STATE_EXITED, PROCESS_STATE_FATAL

Here's a small improvement:

• Use -SIGQUIT rather than hard-coding the signal name
• Use $PPID to get the PID of supervisord (the parent of the event listener)

 while read line; do
   echo "Processing Event: $line" >&2;
-  kill -3 $(cat "/var/run/supervisord.pid")
+  kill -SIGQUIT $PPID
 done < /dev/stdin

Thanks, your save my soul

For those who don't want a separate file 😄

[supervisord]
loglevel=warn
nodaemon=true

[program:hi]
command=bash -c "echo waiting 5 seconds . . . && sleep 5"
autorestart=false
numprocs=1
startsecs=0
stderr_logfile=/dev/stderr
stderr_logfile_maxbytes=0
stdout_logfile=/dev/stdout
stdout_logfile_maxbytes=0

[eventlistener:processes]
command=bash -c "printf 'READY\n' && while read line; do kill -SIGQUIT $PPID; done < /dev/stdin"
events=PROCESS_STATE_STOPPED,PROCESS_STATE_EXITED,PROCESS_STATE_FATAL

I've found the approach mentioned in this comment much more simple.
Supervisor/supervisor#712 (comment)