Description
Currently all timeouts for prometheus calls are 5 sec and not configurable (hardcoded). Sometimes due to environment (or particular server configuration) some calls to prometheus might take more than 5 seconds (for example /-/reload). In order to work around this kind of issues it might be useful to be able to increase those timeouts.
Seems like per-endpoint timeouts will look better than one global configurable prometheus timeout.
Attachments
For Gerrit Dashboard: MB-51619 | ||||||
---|---|---|---|---|---|---|
# | Subject | Branch | Project | Status | CR | V |
173374,7 | MB-51619 Configurable timeouts for prometheus requests | master | ns_server | Status: MERGED | +2 | +1 |
173650,1 | MB-51619 Additional configurable timeouts | master | ns_server | Status: ABANDONED | 0 | +1 |