Table of Contents

• How-to
  • Test a cluster correctly configured
• Reference
• Topics

How-to

Test a cluster correctly configured

• Use statistics API to test the current cluster's status

# For each node
$ curl -s http://127.0.0.1:2379/v2/stats/self | jq

{
  "name": "etcd1",
  "id": "912193dcb3a37035",
  "state": "StateFollower",
  "startTime": "2017-07-17T09:36:46.406539193Z",
  "leaderInfo": {
    "leader": "c7a7eaa884a2af08",
    "uptime": "7m38.68272438s",
    "startTime": "2017-07-17T09:36:46.428994302Z"
  },
  "recvAppendRequestCnt": 2221,
  "recvPkgRate": 4.8195062183526085,
  "recvBandwidthRate": 357.34228855975414,
  "sendAppendRequestCnt": 0
}

# For the whole cluster
curl -s http://127.0.0.1:2379/v2/stats/leader

{
  "leader": "c7a7eaa884a2af08",
  "followers": {
    "912193dcb3a37035": {
      "latency": {
        "current": 0.003626,
        "average": 0.004939527118043262,
        "standardDeviation": 0.003546182920897243,
        "minimum": 0.001497,
        "maximum": 0.069869
      },
      "counts": {
        "fail": 0,
        "success": 2821
      }
    },
    "ec29e3c44c6a8357": {
      "latency": {
        "current": 0.00536,
        "average": 0.14451230827314546,
        "standardDeviation": 7.688751073246117,
        "minimum": 0.001447,
        "maximum": 424.421438
      },
      "counts": {
        "fail": 8,
        "success": 3046
      }
    }
  }
}

Reference

Topics