units stuck in model after failed deploy

[james beedy]

Tried to deploy a charm that uses juju storage to a lxd on a aws instance, got an error that storage wasn't supported, then the units decided to stick around.

To reproduce, just deploy a charm that uses storage to lxd on aws instance using aws provider and the beta juju snap and you should hit this.

e.g

juju add-machine # gets machine 0
juju deploy cs:~jamesbeedy/elasticsearch-19 --to lxd:0

juju status --format yaml http://paste.ubuntu.com/25854172/

juju status http://paste.ubuntu.com/25854173/

[Ian Booth (wallyworld)]

If a unit goes into an error state, it won't be automatically removed. I don't think remove-unit likes to remove units in error either, and sadly there's no --force option. I think we do need to consider supporting "remove-unit --force" to allow such situations to be clean up.

[Canonical Juju QA Bot (juju-qa-bot)]

This bug has not been updated in 2 years, so we're marking it Low importance. If you believe this is incorrect, please update the importance.