Integrate Prometheus

[yaroslav-gwit]

Prometheus will be installed using pkg and will be responsible to collect Hoster metrics, and all the VMs metrics that are running on a specific node.

This will allow to scale Hoster clusters to enormous sizes, because every node will be responsible for it's own monitoring. Whenever a new VM is created/cireset (in production mode) it will immediately be added to the prometheus.yaml to make sure every single VM is being monitored.

UptimeKuma will simply monitor a RestAPI endpoint /api/v1/health-check and the Prometheus endpoint, and will make sure all the nodes are running correctly.

This issue will be closed once the basic integration is done. I'll then open a new one to integrate all these collected metrics into our WebUI, and another one to create Hoster specific Grafana dashboards.

Update 1:
To access the metrics and alerts behind a firewall (in the private deployment scenarios) we can leverage Thanos in the remote write mode and a centralised AlertManager. This will allow us to provide an on-prem SaaS-like Hoster offering. The only concern is alerting, in case if one of the Prometheus instances goes down.