WIP: add prometheus metrics for services

[detongz]

This commit is a proof of concept, each service need to modify its own Roter init function. In NewRouter function, inits the ResponseTime for its namespace(service). These services will have /metrics API for prometheus to pull metrics from them, instead of push metrics to prometheus. And this is not tested yet.

What this PR does / Why we need it:

Add Prometheus metrics for services. Fixes #issues/755

What is changed and how it works?

1. +/metrics API for prometheus to pull metrics
2. WIP/POC: collect response time for APIs.

Does this PR introduce a user-facing change?

• [√] API change
• [-] database schema change
• [-] behavioral change
• [√] change in non-functional attributes such as efficiency or availability
• [√] fix of a previous issue

[detongz]

This commit is a proof of concept, each service need to modify its own Roter init function. In NewRouter function, inits the ResponseTime for its namespace(service). These services will have /metrics API for prometheus to pull metrics from them, instead of push metrics to prometheus.

[landylee007]

long time no response, closed it for new Pull request.