Bot not receiving termination signals

[Trimatix]

When run outside of docker, SIGINT/SIGTERM appear to work fine. Otherwise, the bot is killed but does not appear to have received the signal.

[Laura7089]

Could be related to this: https://stackoverflow.com/questions/62853875/stopping-python-container-is-slow-sigterm-not-passed-to-python-process. If that's the problem, we can bake tini into the docker container and/or add init: true to the docker-compose file.

[Trimatix]

If that's the problem, we can bake tini into the docker container and/or add init: true to the docker-compose file.

Sadly our issue is different, I have a similar python script set up to detect sigint but it prints nothing
I've also already tried using init: true but no dice :(

[Laura7089]

Okay, I've tinkered a bit:

• Running with tini inside the container or the init: true param in compose doesn't solve the issue
• We're using a direct call in the docker exec form so that isn't the issue
• Adding more explicit debug print statements inside handlers and the handler register function shows that the handlers are registered seemingly without issue but the handlers are never called - I've tried creating a marker file as well in case the logs aren't being caught by docker but no beans
• Running the container with entrypoint sleep 99999, then docker-compose exec bot bash and manually running python main.py and issuing ctrl-C directly also doesn't work

[Laura7089]

When run outside of docker, it SIGINT/SIGTERM appear to work fine.

I'm actually unable to replicate this - it doesn't outside of docker either. Could this be because the handlers are registered in a module that isn't the calling one?