38d3dfb84a
Currently, the APISchedulingService only migrates ongoing audits when it detects the decision-engine changes the state from ACTIVE to FAILED. That makes that, audits assigned to a FAILED decision-engine are not reassigned when the watcher-api service is started after the decision-engine dies. After this patch, the APISchedulingService will also migrate ongoint audits also when a decision-engine is detected as FAILED when the service is started. I'm also moving the audit migration logic to a separate function to improve readability. Closes-Bug: #2127777 Change-Id: Ic3c9d89973ca976dde9f6e9b423671e52b049cdb Signed-off-by: Alfredo Moralejo <amoralej@redhat.com>
13 lines
497 B
YAML
13 lines
497 B
YAML
---
|
|
fixes:
|
|
- |
|
|
Currently, the APISchedulingService does not migrate any ongoing
|
|
continuous audits in a failed decision-engine to an alive one when the
|
|
decision-engine dies while the watcher-api is not running.
|
|
|
|
This patch fixes those cases by migrating the audits found on a failed
|
|
decision-engine when the watcher-api is started in addition to when a
|
|
change in a decision-engine status is detected.
|
|
|
|
For more details: https://bugs.launchpad.net/watcher/+bug/2127777
|