Actions
Bug #22670
open
health test point doesn't honour internal error 500 due to passenger not starting
Status:
New
Priority:
Normal
Assigned To:
-
Category:
-
Target version:
-
Story points:
-
Description
Hi, we were experimenting with having the load balancers react to the health ping (curl -H "Authorization: Bearer XXXXX" https://gshlvapi2.rgw.roche.com/_health/ping) and drop the node when it is not responding with 2XX codes.
After this was setup, I forced a internal error 500 by updating concurrent-ruby to 1.3.5, but the status continued being OK, even if this precluded passenger from starting. We would want this health ping to be more representative of the actual working status of the apis/keepweb
No data to display
Actions