Details
-
Bug
-
Status: Closed
-
Major
-
Resolution: Fixed
-
None
Description
when atleast 1 component of scheduler health check status updates to succeeded:"false", status page stops working with error - "Oops, something went wrong. YuniKorn is not accessible."
Rest Call shows on both 9889 and 9080 port -
{"status_code":503,"message":"Scheduler is not healthy","description":"Scheduler is not healthy"}All other rest calls work fine on both ports.
The 503 most likely causes the code in the web UI to shut down without a possibility to start again.
I do not see any code that would handle/retry the 503. It also does not make sense for me to return that code in this case
Attachments
Attachments
Issue Links
- relates to
-
YUNIKORN-2162 Add unit test for checkHealthStatus
- Closed
- links to