QuartzHealthCheck never recovers after detecting failure

[bencyoung-Fignum]

Describe the bug

The error count can only be incremented but is never cleared if things start working.

Version used

3.3.3

To Reproduce

Start a quartz app with the DB down and then start the DB. The health check always reports an error.

Expected behavior

I'd expect the health check to start reporting success when the DB comes up

[bencyoung-Fignum]

I may be able to raise a PR if that's an acceptable route?

[lahma]

Sure, that would be great!

[bencyoung-Fignum]

At the moment it seems to report unhealthy if there's any job errors whereas I'd expect it to be more of an infrastructural test? I.e. it's ok if jobs are failing but not if there are DB errors? Is there a way to tell the difference? The other option is to just check if the job store is healthy rather than listen for general errors?

[lahma]

I think it would be sufficient to check that scheduler is running and job store operates. A single small query should suffice, even trying to get non-existing item as long as error is what is expected.

[bencyoung-Fignum]

I think this is done.

[bencyoung-Fignum]

I've simplified and tested this now