Create prometheus client

[wangchen615]

This PR is to solve #10 and close #10 .

There are the following major changes:

1. The prometheus.go was added under pkg/metricsprovider with FetchAllHostsMetrics method implemented.
2. The pkg/watcher/watcher.go was revised to support different types of clients to fill in the datasource field.

		switch reflect.TypeOf(w.client).Name() {
		case "promClient":
			go windowWatcher(duration, "Prometheus")
		case "metricsServerClient":
			go windowWatcher(duration, "k8s_metricserver")

3. FetchHostMetrics was commented as Prometheus client does not need to fetch host metric one by one.
4. README.md is updated to include a guide on configuring environment variables.

[wangchen615]

A sample output of the load watcher looks like the following:

> curl -H "Content-Type: application/json" http://127.0.0.1:2020/watcher|jq
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100  1900  100  1900    0     0  1855k      0 --:--:-- --:--:-- --:--:-- 1855k
{
  "timestamp": 1612427148,
  "window": {
    "duration": "15m",
    "start": 1612426248,
    "end": 1612427148
  },
  "source": "Prometheus",
  "data": {
    "NodeMetricsMap": {
      "10.95.102.97": {
        "metrics": [
          {
            "name": "instance:node_cpu:ratio",
            "type": "CPU",
            "operator": "avg_over_time",
            "rollup": "15m",
            "value": 0.19731935228375208
          },
          {
            "name": "instance:node_memory_utilisation:ratio",
            "type": "Memory",
            "operator": "avg_over_time",
            "rollup": "15m",
            "value": 0.08354249717493203
          },
          {
            "name": "instance:node_cpu:ratio",
            "type": "CPU",
            "operator": "stddev_over_time",
            "rollup": "15m",
            "value": 0.00048452241606516646
          },
          {
            "name": "instance:node_memory_utilisation:ratio",
            "type": "Memory",
            "operator": "stddev_over_time",
            "rollup": "15m",
            "value": 0.00013245370097128003
          }
        ],
        "tags": {},
        "metadata": {
          "dataCenter": ""
        }
      },
      "10.95.102.99": {
        "metrics": [
          {
            "name": "instance:node_cpu:ratio",
            "type": "CPU",
            "operator": "avg_over_time",
            "rollup": "15m",
            "value": 0.2048283625730656
          },
          {
            "name": "instance:node_memory_utilisation:ratio",
            "type": "Memory",
            "operator": "avg_over_time",
            "rollup": "15m",
            "value": 0.09212922586804047
          },
          {
            "name": "instance:node_cpu:ratio",
            "type": "CPU",
            "operator": "stddev_over_time",
            "rollup": "15m",
            "value": 0.0005520914622929221
          },
          {
            "name": "instance:node_memory_utilisation:ratio",
            "type": "Memory",
            "operator": "stddev_over_time",
            "rollup": "15m",
            "value": 0.0001700369598327637
          }
        ],
        "tags": {},
        "metadata": {
          "dataCenter": ""
        }
      },
      "10.95.102.90": {
        "metrics": [
          {
            "name": "instance:node_cpu:ratio",
            "type": "CPU",
            "operator": "avg_over_time",
            "rollup": "15m",
            "value": 0.05814883040935881
          },
          {
            "name": "instance:node_memory_utilisation:ratio",
            "type": "Memory",
            "operator": "avg_over_time",
            "rollup": "15m",
            "value": 0.10718011619404219
          },
          {
            "name": "instance:node_cpu:ratio",
            "type": "CPU",
            "operator": "stddev_over_time",
            "rollup": "15m",
            "value": 0.00025611915450171206
          },
          {
            "name": "instance:node_memory_utilisation:ratio",
            "type": "Memory",
            "operator": "stddev_over_time",
            "rollup": "15m",
            "value": 3.1065560195209836e-05
          }
        ],
        "tags": {},
        "metadata": {
          "dataCenter": ""
        }
      }
    }
  }
}

[zorro786]

Please review the comments. Also, my name in the commit message is redundant 😄 - Something like "Add Prometheus client" is fine :)

[zorro786]

Thanks for making the changes @wangchen615. LGTM.