Docker rm fails / is there a better way?

[Eric Miller]

This is more a question, but possibly something that people often run into when troubleshooting, so I thought I'd ask the question here - especially if there is a potential workaround, bug identified, or future enhancement.

When deleting a container due to corruption, with the intent to rebuild it with a kolla-ansible deploy, I perform a "docker stop <container>" followed by a "docker rm <container>". However, the "rm" command often results in:
Error response from daemon: Driver devicemapper failed to remove root filesystem aea77eddf440ef4f0c7854d5ec3a3d5e52814be3703ff8100679e23496cb9671: Device is Busy

The only way to resolve this is to restart Docker (and thus all containers) on the host.

Is there something that can be done about this? Everything I have read so far indicates that overlays need to be used instead of devicemapper mounts? I'm not a Docker expert, so I thought I'd ask here - whether there is a different way to remove a container for re-deployment, or does something need to be done prior to the "rm" command?

Thanks!

Eric