docker start / stop / restart / rm

Manage the lifecycle of existing containers: stop and start them without losing their state, restart them, or remove them for good.

created → running → paused / exited → removed. Stopping keeps the container (and its writable layer) on disk; removing deletes it.

Commands

SHELL

1docker stop <container> # SIGTERM, then SIGKILL after 10s

2docker stop -t 30 <container> # give it 30s to shut down cleanly

3docker start <container> # start a stopped container again

4docker start -ai <container> # start and attach (interactive)

5docker restart <container> # stop + start

6docker kill <container> # immediate SIGKILL

7docker pause / unpause <container> # freeze / resume processes

8docker rm <container> # remove a stopped container

9docker rm -f <container> # stop and remove in one go

10docker rename <old> <new>

Many containers at once

SHELL

1docker stop $(docker ps -q) # stop all running

2docker rm $(docker ps -aq -f status=exited) # remove all exited

3docker container prune # remove all stopped (asks)

4docker update --restart unless-stopped <c> # change restart policy

Policy	Behaviour
`no` (default)	Never restart automatically
`on-failure[:N]`	Restart on non-zero exit, optionally max N times
`always`	Always restart, also after daemon restart
`unless-stopped`	Like always, but not if you stopped it manually

docker rm -f kills without a graceful shutdown; databases may need recovery. Prefer stop first.