As a Prism Server user I want to be able to set log severity level so I can have less log load

[panagiks]

Context

I'm running Prism Server on a k8s environment and best practice is to have health/liveness checks. In the absence of a healthcheck endpoint on the prism server itself I've used one of the 'generated' endpoints as a target. This in turn causes 5 lines of logs / request / instance (3 instances required for minimum HA) and with both checks set to (default) 5 seconds this means 30 lines of logs / 5 seconds (5 lines/check/5seconds/instance).

Over the time this causes a significant log load and makes it harder to process / identify actual errors.

Acceptance Criteria

• [ ] There is either a CLI flag or an environment variable that allows the user to configure the Prism Server log level

Sprint Ready Checklist

• [x] Acceptance criteria defined
• [ ] Team understands acceptance criteria
• [ ] Team has defined solution / steps to satisfy acceptance criteria
• [ ] Acceptance criteria is verifiable / testable
• [ ] External / 3rd Party dependencies identified

[XVincentX]

No changes in the code are necessary.

I suggest to pipe the output to jq and filter out the level you do not want or use pino-pretty and its level flag to get out what you are not interested into.