chore(deps): update prometheus.client

[phnx47]

Fixes #1772

@tomkerkhove I can't remove services.BuildServiceProvider(); without completely rewrite schedule logic. Right now I just updated Prometheus.Client and use DefaultCollectorRegistry.

Next step is rewrite scheduler or we can save it as is... For rewrite we should register Scheduler Job in runtime, Configure section instead of ConfigureServices section, but I never used CronScheduler.AspNetCore. I found this kdcllc/CronScheduler.AspNetCore#29, maybe you know how to register in runtime better.

[tomkerkhove]

Thanks for jumping in! Would you mind fixing the CI please?

[phnx47]

CI - Code / Verify Codebase (pull_request): Seems this error - dorny/test-reporter#168, but permissions already added.

I ran it in my fork and all tests passed:
https://github.com/phnx47/promitor/actions/runs/3183407827/jobs/5190580488
https://github.com/phnx47/promitor/runs/8703462355

So, something weird, I can't re-run it here...

[phnx47]

Promitor CI - Resource Discovery Agent and Promitor CI - Scraper Agent: The job running on agent Azure Pipelines 3 ran longer than the maximum time of 60 minutes

Step: Run Integration Tests

Starting test execution, please wait...
A total of 1 test files matched the specified pattern.
Current metric type is not a gauge, ignoring.
...
Current metric type is not a gauge, ignoring.
##[error]The operation was canceled.

[tomkerkhove]

CI - Code / Verify Codebase (pull_request): Seems this error - dorny/test-reporter#168, but permissions already added.

I ran it in my fork and all tests passed: phnx47/promitor/actions/runs/3183407827/jobs/5190580488 phnx47/promitor/runs/8703462355

So, something weird, I can't re-run it here...

Re-triggered, let's see as it's odd indeed.

[tomkerkhove]

Promitor CI - Resource Discovery Agent and Promitor CI - Scraper Agent: The job running on agent Azure Pipelines 3 ran longer than the maximum time of 60 minutes

Step: Run Integration Tests

Starting test execution, please wait...
A total of 1 test files matched the specified pattern.
Current metric type is not a gauge, ignoring.
...
Current metric type is not a gauge, ignoring.
##[error]The operation was canceled.

Yes, I've just noticed so that must mean something is wrong with publishing the metrics. Quickly did #2144, let's see if it works.

[tomkerkhove]

LGTM, thanks!

[phnx47]

@tomkerkhove What about rewrite Scheduler to register jobs in runtime? I Can find time for this. It is not only for prometheus client https://learn.microsoft.com/en-us/dotnet/core/extensions/dependency-injection-guidelines#recommendations

[tomkerkhove]

Good callout - I forgot to open an issue for that. Would you mind doing that with some context please?

[phnx47]

@tomkerkhove Sure, I will do.