把prometheus和golang结合在一起，并监控golang的性能和接口状态。

1. golang的使用

1.1 基础的内置指标

下载包

1
2
3

go get github.com/prometheus/client_golang/prometheus
go get github.com/prometheus/client_golang/prometheus/promauto
go get github.com/prometheus/client_golang/prometheus/promhttp

代码注册

package main

import (
        "net/http"

        "github.com/prometheus/client_golang/prometheus/promhttp"
)

func main() {
        http.Handle("/metrics", promhttp.Handler())
        http.ListenAndServe(":2112", nil)
}

1	go run main.go

访问获取信息

1	curl http://localhost:2112/metrics

配置prometheus

1	sudo vi /opt/prometheus/prometheus.yml

- job_name: 'golang'
  scrape_interval: 10s
  static_configs:
    - targets: ['127.0.0.1:2112']

1	sudo systemctl restart prometheus

导入模板 6671

1.2 gin 包装 prometheus

下载包

1	go get github.com/zsais/go-gin-prometheus

代码

// prometheus wrap
p := ginprometheus.NewPrometheus("gin")
p.ReqCntURLLabelMappingFn = func(c *gin.Context) string {
	url := c.Request.URL.Path
	for _, p := range c.Params {
		if p.Key == "id" {
			url = strings.Replace(url, p.Value, ":id", 1)
			break
		} else if p.Key == "short_link" {
			url = strings.Replace(url, p.Value, ":short_link", 1)
			break
		}
	}
	return url
}
p.Use(appEngine)

注意这段代码尽量靠前放，在其他中间件之前。

默认地址是/metrics

http://127.0.0.1:9061/metrics

模板

# 接口请求
gin_requests_total{job="golang"}       
Legend： {{url}}  


#错误码
promhttp_metric_handler_requests_total{job="golang", instance="127.0.0.1:9061"}
Legend：{{code}}

1.3 另外一个库

https://github.com/penglongli/gin-metrics

2. 报警机制

2.1 Alert监控报警

打开配置

sudo vi /etc/grafana/grafana.ini 
#In your custom configuration file ($WORKING_DIR/conf/custom.ini), go to the unified alerts section. Set the enabled property to true.

sudo systemctl restart grafana-server.service

新建报警API

飞书不支持，自己用接口中转一下吧。

报警匹配规则

3. 遇到的问题

3.1 删除old jobs和instance

Expanding on evgenyl’s answer, the exact command would be something like:

1	curl -X POST -g 'http://localhost:9090/api/v1/admin/tsdb/delete_series?match[]={job="name_of_old_job"}'

Replace name_of_old_job with the name of the job you want deleted.

Reminder that you need to have started prometheus with the –web.enable-admin-api flag

1
2

curl -X POST -g 'http://localhost:9090/api/v1/admin/tsdb/delete_series?match[]={job="linux-node"}'
curl -X POST -g 'http://localhost:9090/api/v1/admin/tsdb/delete_series?match[]={instance="localhost:9100"}'

不过需要注意的是上面的 API 调用并不会立即删除数据，实际数据任然还存在磁盘上，会在后面进行数据清理。

Levon's Blog

prometheus监控golang