GoFrame 框架:添加 Prometheus 监控

2022年01月14日 阅读数:5
这篇文章主要向大家介绍GoFrame 框架:添加 Prometheus 监控,主要内容包括基础应用、实用技巧、原理机制等方面,希望对大家有所帮助。

介绍

经过一个完整例子,在基于 GoFrame 框架的微服务中添加 Prometheus 监控。git

GoFrame 框架监控中间件,会在后续的文章中介绍。github

咱们将会使用 rk-boot 来启动 GoFrame 微服务。golang

请访问以下地址获取完整教程:docker

安装

go get github.com/rookie-ninja/rk-boot
go get github.com/rookie-ninja/rk-gf

快速开始

1.建立 boot.yaml

boot.yaml 文件描述了 GoFrame 框架启动参数,rk-boot 经过读取 boot.yaml 来启动 GoFrame。shell

---
gf:
  - name: greeter
    port: 8080
    enabled: true
    prom:
      enabled: true      # Enable prometheus client
#      path: "metrics"   # Default value is "metrics", set path as needed.

2.建立 main.go

// Copyright (c) 2021 rookie-ninja
//
// Use of this source code is governed by an Apache-style
// license that can be found in the LICENSE file.
package main

import (
	"context"
	"github.com/rookie-ninja/rk-boot"
	_ "github.com/rookie-ninja/rk-gf/boot"
)

// Application entrance.
func main() {
	// Create a new boot instance.
	boot := rkboot.NewBoot()

	// Bootstrap
	boot.Bootstrap(context.Background())

	// Wait for shutdown sig
	boot.WaitForShutdownSig(context.Background())
}

3.启动 main.go

$ go run main.go

4.验证

访问: http://localhost:8080/metricsapp

Prometheus 客户端中添加监控

咱们须要先了解 Prometheus 中的以下概念。框架

名字 详情
MetricsSet RK 自定义的结构,经过 MetricsSet 注册 Prometheus 的 Counter,Gauge,Histogram 和 Summary
Prometheus Registerer Prometheus 会经过 Registrerer 来管理 Counter,Gauge,Histogram 和 Summary
Prometheus Counter Counter 是一个累积度量,表示单个单调增长的计数器,其值只能增长或重置为零
Prometheus Gauge Gauge 值能够随意加减
Prometheus Histogram Histogram 进行采样(一般是请求持续时间或响应大小之类的内容)并将它们计算在可配置的桶中,同时还提供全部观测值的总和
Prometheus Summary 与 Histogram 相似,摘要样本观察(一般是请求持续时间和响应大小之类的东西)
Prometheus Namespace Prometheus 监控名格式: namespace_subSystem_metricsName
Prometheus SubSystem Prometheus 监控名格式: namespace_subSystem_metricsName

1.在 main.go 中添加监控项

// Copyright (c) 2021 rookie-ninja
//
// Use of this source code is governed by an Apache-style
// license that can be found in the LICENSE file.
package main

import (
	"context"
	"github.com/rookie-ninja/rk-boot"
	"github.com/rookie-ninja/rk-gf/boot"
	"github.com/rookie-ninja/rk-prom"
)

// Application entrance.
func main() {
	// Create a new boot instance.
	boot := rkboot.NewBoot()

	// Bootstrap
	boot.Bootstrap(context.Background())

	// Create a metrics set into prometheus.Registerer
	set := rkprom.NewMetricsSet("rk", "demo",boot.GetEntry("greeter").(*rkgf.GfEntry).PromEntry.Registerer)

	// Register counter, gauge, histogram, summary
	set.RegisterCounter("my_counter", "label")
	set.RegisterGauge("my_gauge", "label")
	set.RegisterHistogram("my_histogram", []float64{}, "label")
	set.RegisterSummary("my_summary", rkprom.SummaryObjectives, "label")

	// Increase counter, gauge, histogram, summary with label value
	set.GetCounterWithValues("my_counter", "value").Inc()
	set.GetGaugeWithValues("my_gauge", "value").Add(1.0)
	set.GetHistogramWithValues("my_histogram", "value").Observe(0.1)
	set.GetSummaryWithValues("my_summary", "value").Observe(0.1)
	
	// Wait for shutdown sig
	boot.WaitForShutdownSig(context.Background())
}

2.启动 main.go

$ go run main.go

3.验证

访问: http://localhost:8080/metricside

推送到 pushgateway

接下来,咱们看一下,如何自动把监控数据推送到远程 pushgateway 中。微服务

1.boot.yaml 中启动 pusher

---
gf:
  - name: greeter
    port: 8080
    enabled: true
    prom:
      enabled: true                         # Enable prometheus client
      pusher:
        enabled : true                      # Enable backend job push metrics to remote pushgateway
        jobName: "demo"                     # Name of current push job
        remoteAddress: "localhost:9091"     # Remote address of pushgateway
        intervalMs: 2000                    # Push interval in milliseconds
#        basicAuth: "user:pass"             # Basic auth of pushgateway
#        cert:
#          ref: "ref"                       # Cert reference defined in CertEntry. Please see advanced user guide for details.

2.在本地启动 pushgateway

咱们使用 docker 启动 pushgatewayui

$ docker run -p 9091:9091 prom/pushgateway

3.启动 main.go

$ go run main.go

4.验证

访问: http://localhost:9091/metrics

完整选项

名字 描述 类型 默认值
gf.prom.enabled 启动 prometheus boolean false
gf.prom.path Prometheus Web 路径 string /metrics
gf.prom.pusher.enabled 启动 prometheus pusher bool false
gf.prom.pusher.jobName JobName 将会以标签的形式添加到监控指标,并推送到远程 pushgateway string ""
gf.prom.pusher.remoteAddress Pushgateway 远程地址, http://x.x.x.x 或者 x.x.x.x string ""
gf.prom.pusher.intervalMs 推送间隔(毫秒) string 1000
gf.prom.pusher.basicAuth 远程 Pushgateway 的 Basic auth。 格式:[user:pass] string ""
gf.prom.pusher.cert.ref rkentry.CertEntry 的引用,请参考官网高级指南 string ""