"docker stop factorio" does not allow enough time for large saves to properly save progress

[rhwas]

I have a large save file (~50 MB) and when running the command "docker stop factorio", the save data is not properly saved. (see docker logs below)

Even adding a timeout does not solve the problem (docker stop -t 120 factorio) and actually does not seem to wait 120seconds at all. The only result is a new MainServer.tmp.zip in the ./saves folder which is not complete.

"""
668.651 Received SIGTERM, shutting down
668.651 Quitting: signal.
668.651 Info MainLoop.cpp:437: Saving map as /factorio/saves/MainServer.zip
668.896 Info MainLoop.cpp:448: Saving progress: 0.003471%
669.896 Info MainLoop.cpp:448: Saving progress: 55.381626%
Session terminated, killing shell... ...killed.

[SuperSandro2000]

Did you try a timeout of 600?

[rhwas]

I have tried higher timeout values, however, it seems to not even care and immediately interrupts without a timeout.

Looking deeper,

Because PID 1 is runuser SIGTERM goes to runuser.
runuser forwards SIGTERM to Factorio,
Factorio begins saving but then exits early,
runuser exits immediately when Factorio exits,
Container stops → saving is interrupted

What does work is to run "docker exec factorio rcon /quit", and then, sleep for like 60 seconds, and then run "docker stop factorio"

[dtandersen]

Does docker stop --signal SIGINT work?