Minor tweaks to imaging support:
1. Use distinct imaging-failed status to indicate a failure, instead of leaving instance in imaging. 2. Leave the refresh button enabled so that user can refresh the instance, and if the backend sliver is "ready" then switch the instance back to ready. 3. The snapshot loop now does the same as above; tell the user the imaging failed, but move the instance back to ready if the sliver is ready.
Showing with 29 additions and 7 deletions