Containers using dinit core-dump on stop because of STOPSIGNAL SIGQUIT

The Basics

  • Installation method (e.g. AlO, NCP, Bare Metal/Archive, etc.)
    • AIO

Since the switch from supervisor to dinit replace supervisor with dinit by Zoey2936 · Pull Request #8533 · nextcloud/all-in-one · GitHub stopping the master- or the nextcloud container kills it with a core dump. The dump also gets printed in any active ssh sessions on the host. (That’s how I found out about it)

    Process 1680569 (dinit) of user 0 dumped core.
    Stack trace of thread 7:
    #0  0x00007fdce1b6b8f3 n/a (/lib/ld-musl-x86_64.so.1 + 0x698f3)
Command Line: dinit --system --container php-fpm cron taskprocessing-worker run-exec-commands is-nextcloud-online
    Hostname: nextcloud-aio-nextcloud
      Signal: 3 (QUIT)
   Timestamp: Wed 2026-10-14 15:21:54 CEST

Cause

Both images set STOPSIGNAL SIGQUIT. dinit only handles SIGQUIT if it runs as PID 1 but it does run under docker-init which forwards the signal. The default handler for SIGQUIT is to exit with a core dump.

Fix

instead of using SIGQUIT use SIGTERM to cleanly shutdown all the services before exiting.

was adressed by #8747 in already available AIO v14.3.0.
pls update