device stop doesn't handle restarting (crashing) containers
adamkdean opened this issue · comments
TL;DR I think this is misreporting by CLI due to limits of the old container matching behaviour. As of 9cf7413 this problem no longer exists.
--
This service would have been started with SPECIFIC_TAG=0.1.0-10 edgetest device start
meaning that subsequent interactions with the container would also require SPECIFIC_TAG
to be set, otherwise there would be no match for the container based on image name. Ergo, in the screenshot the edgetest device stop
looked for a container running :latest
, found none, and reported as such.
Now that we are using container name for matching basis, the container should always be matched if it exists, even if it's stuck in a restart loop. This will be particularly useful for update
et al. where the image name is likely to vary between versions.