Improve the feedback when NoValidHost is raised by the Scheduler
Currently when the scheduler cannot find a host to place a newly-requested instance, there is only partial information about why the placement failed, and this is scattered among various logs, making it difficult to find. Improving this situation has been identified as a priority task for the Liberty release.
Note: there was some prior work related to this (https:/
Blueprint information
- Status:
- Complete
- Approver:
- John Garbutt
- Priority:
- Medium
- Drafter:
- Ed Leafe
- Direction:
- Approved
- Assignee:
- Ed Leafe
- Definition:
- Approved
- Series goal:
- Accepted for liberty
- Implementation:
- Implemented
- Milestone target:
- 12.0.0
- Started by
- John Garbutt
- Completed by
- John Garbutt
Related branches
Related bugs
Bug #1431291: Scheduler Failures are no longer logged with enough detail for a site admin to do problem determination | Invalid |
Sprints
Whiteboard
Gerrit topic: https:/
Addressed by: https:/
Improve reporting when Scheduler returns NoValidHost
Addressed by: https:/
Add logging when filtering returns nothing.
Addressed by: https:/
Add a Distinct Exception for Exceeding Max Retries
Addressed by: https:/
Make NoValidHost exceptions clearer