Stop dependent containers together with AIO

[baptit]

If administrators do not realize that Nextcloud containers continue to run after AIO shutdown, they are stopped ungracefully in the course of other administrative tasks such as dockerd termination or system reboot because these tasks do not give Nextcloud containers enough time to shut down.

It would be much better for AIO to gracefully terminate the dependent containers as part of its own shutdown. In the course of an AIO update this may not be desirable, but surely for most manually triggered shutdowns.

https://github.com/nextcloud/all-in-one#how-to-stopstartupdate-containers-or-trigger-the-daily-backup-from-a-script-externally is a workaround only for those who have realized the problem, and then it is an additional step in all workflows (scripts, crontab, etc.).

[nihebe]

+1 for this – as pointed out here, the automation of a proper shutdown is pretty complicated when it comes to shutdown events (e.g. USV-triggered shutdown). It would be really desirable to have AIO fully support a regular sigterm event on Docker daemon.

[szaimen]

Hi, I fear your suggestion is not going to fix the underlying problem you are running into.

If administrators do not realize that Nextcloud containers continue to run after AIO shutdown, they are stopped ungracefully in the course of other administrative tasks such as dockerd termination or system reboot because these tasks do not give Nextcloud containers enough time to shut down.

This can be fixed on your system by following the following instructions:

• Increase Docker Shutdown Timeout in Systemd #8324