In a support thread titled “Can’t stop container,” freemangl reports being unable to stop a Docker container using docker stop, even after restarting the Docker service and physical server. They reference a known issue with runsv hanging during shutdown, noting that a fix is available in a newer image — but they can’t rebuild until the old container is stopped. fefrei suggests rebooting the machine, identifying the container via docker ps, and forcefully killing it with docker kill <id>, noting that killing works even when stopping fails. However, freemangl reports that docker kill also hangs indefinitely. fefrei expresses uncertainty, cautioning that killing only works if the container is running normally — and that failed stop attempts may leave the container in an unresponsive, high-CPU state. Ultimately, freemangl confirms that after restarting the physical server, the issue resolved itself and the container could be stopped successfully. The key takeaway: a full server reboot was necessary to restore container control before rebuilding with the updated image.
``
Stopping old container
I just find this:
Continuing the discussion from Runsv hanging on Docker container shutdown:
but, if I can’t stop old container how can I rebuild with the new image?
Reboot your machine so Discourse is running normally.
Run docker ps to get the container ID. If there are multiple, you need to find the right one. Then, run docker kill <id> to kill the container. While stopping is broken, killing works – so violence does help here 
Then, simply rebuild as usual.
thanks, but,
docker kill also takes forever… 
Hm, then I’m out of ideas.
Are you certain you rebooted your server beforehand? Any try to stop the container will most likely leave it unstoppable and needing 100% CPU. Killing only worked for me while the container was running normally.
thanks, I restarted my Physical Server, than worked!