Add custom Prometheus metrics for CPU allocation monitoring

[AutuSnow]

We are currently in the stage of improving functionality, so I think this issue may also be within version 0.1.0. Currently, only the default Go runtime metrics are exposed in the/metrics endpoint, without any business metrics. I think we need to improve this feature point, but I don't think we need to add many metrics in the initial stage. We only need to add some key metrics. Of course, we may need to discuss which metrics we need to expose

[ffromani]

yes, observability metrics is totally something we want in the near future

[AutuSnow]

I want to enter this job in version 0.1.0 because it does not require adding most of the time investment

[AutuSnow]

/assign