[RFC] Diversify and standardize Sentinel metrics and monitoring

## Issue Description

Type: *feature request*

### Describe what feature you want

Currently the metrics and monitoring of Sentinel is not standardized and versatile enough. This issue will discuss about standardization and improvements of metrics and monitoring of Sentinel, so that we can enable:

- provide standard and diverse metrics representation (e.g. percentile, gauge, histogram)
- standardization of monitoring API and transport server API, and adapt to open-source monitoring components (e.g. Prometheus)
- integration with open-source visualization solution (e.g. Grafana) in Sentinel Dashboard

Suggestions are welcomed :)

The issue names no files, tests, or specific entry points. Start by examining Sentinel's existing metrics, monitoring API, transport server API, and Sentinel Dashboard; done would require an agreed scope for standardized percentile, gauge, and histogram representations plus monitoring, Prometheus, and Grafana integration.

Written by the indexing model from the issue text.