Unable to start workspace. Stuck at stopping.

[pybern]

Bug description

Attempts to start a workspace goes from preparing to stopping and gets stuck there. The workspace was last used a couple of hours ago.

Steps to reproduce

Open workspace.

Workspace affected

https://pybern-dream-ujgolxy8don.ws-us64.gitpod.io/

Expected behavior

No response

Example repository

No response

Anything else?

No response

[pybern]

#12939 linking similar issue

[axonasif]

Related: #12908

[geropl]

@kylos101 I think there are two or three parts in here:

1. Why can't the workspace start properly?
2. Why does bridge receive updates for stopping, but never sees any stopped state?
3. Why does bridge not clean those cases up, eventually? (being worked on in [bridge] Logs stuck workspace instances to validate fix #12902 )

Regarding:

1. Sounds like an effect of Workspaces do not initialize because they cannot connect to ws-daemon  #12908. D'accord?
2. Is Team Workspace looking into this? We could help with providing traces for a single workspaces as an example 🤔 Or: Did this happen because we de-registered the cluster before all workspaces were in stopped case? Then we started working on it already ([ws-manager-bridge] Garbage-collect workspace instances whose workspace clusters are not available anymore #6770).

[geropl]

I assume this is a dupe of #12902 and #6770 right now. Will close this as those have already been scheduled.