溫馨提示×
您好，登錄后才能下訂單哦！
點擊 登錄注冊 即表示同意《億速云用戶服務條款》
您好，登錄后才能下訂單哦！
這期內容當中小編將會給大家帶來有關go監控方案的實現是怎樣的，文章內容豐富且以專業的角度為大家分析和敘述，閱讀完這篇文章希望大家可以有所收獲。
數據采集使用go-metrics
傳輸使用UDP, 仿StatsD上傳采集數據, InfluxDB進行數據存儲, Grafana進行展示。
https://github.com/solate/metrics
該地址有已經改好的配置文件可以直接使用。
使用的all-in-one :
git docker-statsd-influxdb-grafana
docker hub 地址
//掛載配置文件，已修改statsd模版 docker run --ulimit nofile=66000:66000 -v /root/telegraf.conf:/etc/telegraf/telegraf.conf -d --name docker-statsd-influxdb-grafana -p 3003:3003 -p 3004:8888 -p 8086:8086 -p 8125:8125/udp samuelebistoletti/docker-statsd-influxdb-grafana:latest
register 使用的name 必須是不同的
將 [[inputs.statsd]] 部分配置打開, 修改templates為:
templates = [ "* measurement.measurement.field" ]
表示傳值prefix.name.field 最好表示為 prefix_name field
package client
import (
"bufio"
"bytes"
"github.com/rcrowley/go-metrics"
"log"
"net"
"strconv"
"strings"
"time"
)
const (
// UDP packet limit, see
// https://en.wikipedia.org/wiki/User_Datagram_Protocol#Packet_structure
UDP_MAX_PACKET_SIZE int = 64 * 1024
)
// Config provides a container with configuration parameters for
// the StatsD exporter
type Config struct {
Network string // Network: tcp, udp.
Addr string // Network address to connect to | 地址
Registry metrics.Registry // Registry to be exported | metrics注冊
FlushInterval time.Duration // Flush interval | 刷新間隔時間
Prefix string // Prefix to be prepended to metric names | 前綴名字
Rate float32 // Rate
Tags string // tag //TODO
conn net.Conn
}
func StatsD(r metrics.Registry, d time.Duration, prefix string, network string, addr string, rate float32) {
conn, err := net.Dial(network, addr)
if err != nil {
panic("conn remote err!")
}
StatsDWithConfig(Config{
Network: network,
Addr: addr,
Registry: r,
FlushInterval: d,
Prefix: prefix,
Rate: rate,
conn: conn,
})
}
// WithConfig is a blocking exporter function
func StatsDWithConfig(c Config) {
for _ = range time.Tick(c.FlushInterval) {
if err := statsd(&c); err != nil {
log.Println(err)
c.conn.Close()
}
}
}
func statsd(c *Config) (err error) {
w := bufio.NewWriter(c.conn)
c.Registry.Each(func(name string, i interface{}) {
switch metric := i.(type) {
case metrics.Counter:
ms := metric.Snapshot()
w.Write(statsdLine(c.Prefix, name, "", ms.Count(), "|c", c.Rate))
case metrics.Gauge:
ms := metric.Snapshot()
w.Write(statsdLine(c.Prefix, name, "", ms.Value(), "|g", c.Rate))
case metrics.GaugeFloat64:
ms := metric.Snapshot()
w.Write(statsdLine(c.Prefix, name, "", ms.Value(), "|g", c.Rate))
case metrics.Histogram:
ms := metric.Snapshot()
ps := ms.Percentiles([]float64{0.5, 0.75, 0.95, 0.99, 0.999, 0.9999})
fields := make([][]byte, 12)
fields = append(fields, statsdLine(c.Prefix, name, "count", ms.Count(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "max", ms.Max(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "mean", ms.Mean(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "min", ms.Min(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "stddev", ms.StdDev(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "variance", ms.Variance(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "p50", ps[0], "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "p75", ps[1], "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "p95", ps[2], "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "p99", ps[3], "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "p999", ps[4], "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "p9999", ps[5], "|g", c.Rate))
buf := bytes.Join(fields, []byte{})
w.Write(buf)
case metrics.Meter:
ms := metric.Snapshot()
fields := make([][]byte, 5)
fields = append(fields, statsdLine(c.Prefix, name, "count", ms.Count(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "m1", ms.Rate1(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "m5", ms.Rate5(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "m15", ms.Rate15(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "mean", ms.RateMean(), "|g", c.Rate))
buf := bytes.Join(fields, []byte{})
w.Write(buf)
case metrics.Timer:
ms := metric.Snapshot()
ps := ms.Percentiles([]float64{0.5, 0.75, 0.95, 0.99, 0.999, 0.9999})
fields := make([][]byte, 12)
fields = append(fields, statsdLine(c.Prefix, name, "count", ms.Count(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "max", ms.Max(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "mean", ms.Mean(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "min", ms.Min(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "stddev", ms.StdDev(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "variance", ms.Variance(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "p50", ps[0], "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "p75", ps[1], "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "p95", ps[2], "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "p99", ps[3], "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "p999", ps[4], "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "p9999", ps[5], "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "m1", ms.Rate1(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "m5", ms.Rate5(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "m15", ms.Rate15(), "|g", c.Rate))
fields = append(fields, statsdLine(c.Prefix, name, "mean", ms.RateMean(), "|g", c.Rate))
buf := bytes.Join(fields, []byte{})
w.Write(buf)
//case metrics.Healthcheck:
// metric.Check()
// log.Printf("healthcheck %s\n", name)
// log.Printf(" error: %v\n", metric.Error())
//
//case metrics.EWMA:
//case metrics.Sample:
}
w.Flush()
})
return
}
//構造發送line
func statsdLine(prefix, name, field string, value interface{}, suffix string, rate float32) []byte {
//<metricname>:<value>|<type>|@<rate>
var buffer bytes.Buffer
//buf := make([]byte, UDP_MAX_PACKET_SIZE)
//添加前綴
if prefix != "" {
//buf = append(buf, prefix...)
//buf = append(buf, '.')
buffer.WriteString(prefix)
buffer.WriteString(".")
} else {
buffer.WriteString("statsd")
buffer.WriteString(".")
}
////添加名稱
//buf = append(buf, name...)
//buf = append(buf, ':')
//將name注冊中的'.'替換成'_', 配合telegraf修改模版,防止將數據庫名字改為屬性
if strings.Contains(name, ".") {
name = strings.ReplaceAll(name, ".", "_")
}
//添加名稱
buffer.WriteString(name)
if field != "" {
buffer.WriteString(".")
buffer.WriteString(field)
}
buffer.WriteString(":")
buf := buffer.Bytes()
switch v := value.(type) {
case string:
buf = append(buf, v...)
case int64:
buf = strconv.AppendInt(buf, v, 10)
case float64:
buf = strconv.AppendFloat(buf, v, 'f', -1, 64)
default:
return nil
}
if suffix != "" {
buf = append(buf, suffix...)
}
if rate != 0 && rate < 1 {
buf = append(buf, "|@"...)
buf = strconv.AppendFloat(buf, float64(rate), 'f', 6, 32)
}
buf = append(buf, "\n"...) //每一行打一個回車，telegraf 是使用回車進行讀取的
return buf
}上述就是小編為大家分享的go監控方案的實現是怎樣的了，如果剛好有類似的疑惑，不妨參照上述分析進行理解。如果想知道更多相關知識，歡迎關注億速云行業資訊頻道。
免責聲明：本站發布的內容（圖片、視頻和文字）以原創、轉載和分享為主，文章觀點不代表本網站立場，如果涉及侵權請聯系站長郵箱：is@yisu.com進行舉報，并提供相關證據，一經查實，將立刻刪除涉嫌侵權內容。
