方框_golangpromethusconsul服务发现

篇首语：本文由编程笔记#小编为大家整理，主要介绍了golang promethus consul 服务发现相关的知识，希望对你有一定的参考价值。

docker pull consul
docker pull prom/prometheus
docker pull grafana/grafana


brew方式&＃xff1a;

brew install grafana # grafana
brew tap hashicorp/tap # consul
brew install hashicorp/tap/consul # consul
brew install prometheus # prometheus 安装有点问题, 作者采用了docker方式


启动grafana&＃xff1a;

brew services start grafana


访问&＃xff1a;http://127.0.0.1:3000/

登录的时候默认用户名密码均为&＃xff1a;admin

启动promethus:

docker run -d -p 8091:9090 --name&＃61;prometheus_server prom/prometheus # 主机9090被占用, 临时换成了8091


成功后访问http://127.0.0.1:8091/classic/graph

consul 前台启动&＃xff1a;

consul agent -dev


这种方式只需要 ctrl &＃43; c 即可关闭 consul

正常启动&＃xff1a;

consul agent -server -ui -bootstrap-expect&＃61;1 -data-dir&＃61;/Users/xxxxxx/consul_log -node&＃61;server-1 -client&＃61;0.0.0.0 -bind&＃61;192.168.3.241 -datacenter&＃61;dc1


启动成功后访问8500端口会出现如下界面:

自此所有组件启动完成(失败的情况请大家自行Google)

Promethus 配置pull方式

docker exec -it eb6c1878a8dc /bin/sh # eb6c1878a8dc 为promethus容器的id
vi /etc/prometheus/prometheus.yml


添加服务发现相关配置&＃xff1a;

scrape_configs:
# - job_name: "prometheus"
# static_configs:
# - targets: ["192.168.3.241:22200","192.168.3.241:22201"]
- job_name: &＃39;prometheus_monitor&＃39;
consul_sd_configs:
- server: &＃39;192.168.3.241:8500&＃39;
services: [prometheus_monitor]


static_configs中的targets为固定拉取部分&＃xff0c;在容器化的今天&＃xff0c;pod扩展后都需要手动修改很麻烦&＃xff0c;采用服务发现的方式来发现节点才是理想的解决方式。services 是程序注册的服务名称&＃xff0c;也支持正则的方式&＃xff0c;具体请大家自行Google一下。

改完后的配置如下&＃xff1a;

请确保server是可以访问到的地址&＃xff0c;包括局域网或公网能否访问consul地址

客户端代码:

package main
import (
"flag"
"fmt"
"github.com/go-basic/uuid"
consulapi "github.com/hashicorp/consul/api"
"github.com/prometheus/client_golang/prometheus"
"github.com/prometheus/client_golang/prometheus/promauto"
"github.com/prometheus/client_golang/prometheus/promhttp"
"net"
"net/http"
"time"
)
var (
commandPort int
instanceId string // 实例Id
opsProcessed prometheus.Counter
)
const (
consulAddress &＃61; "127.0.0.1:8500"
)
func RecordMetrics()
for
opsProcessed.Inc()
time.Sleep(5 * time.Second)


func GetLocalIP() string
addressList, err :&＃61; net.InterfaceAddrs()
if err !&＃61; nil
return ""

for _, address :&＃61; range addressList
if ipNet, ok :&＃61; address.(*net.IPNet); ok && !ipNet.IP.IsLoopback()
if ipNet.IP.To4() !&＃61; nil
return ipNet.IP.String()



return ""

func main()
flag.IntVar(&commandPort, "command_port", 22200, "command port")
flag.Parse()
fmt.Println("command_port: ", commandPort)
RegisterService()
opsProcessed &＃61; promauto.NewCounter(prometheus.CounterOpts
Name: "myapp_processed_ops_total",
Help: "The total number of processed events",
ConstLabels: map[string]string
"instance_id": instanceId,
,
)
go RecordMetrics()
http.HandleFunc("/", Handler)
http.Handle("/metrics", promhttp.Handler())
err :&＃61; http.ListenAndServe(fmt.Sprintf(":%d", commandPort), nil)
if err !&＃61; nil
fmt.Println("error: ", err.Error())


func RegisterService()
// 创建连接consul服务配置
config :&＃61; consulapi.DefaultConfig()
config.Address &＃61; consulAddress
client, err :&＃61; consulapi.NewClient(config)
if err !&＃61; nil
fmt.Println("consul client error : ", err)
return

// 创建注册到consul的服务到
id :&＃61; uuid.New()
instanceId &＃61; id
registration :&＃61; new(consulapi.AgentServiceRegistration)
registration.ID &＃61; id
registration.Name &＃61; "prometheus_monitor" //根据这个名称来找这个服务
registration.Port &＃61; commandPort
registration.Address &＃61; GetLocalIP()
// 增加consul健康检查回调函数
check :&＃61; new(consulapi.AgentServiceCheck)
check.HTTP &＃61; fmt.Sprintf("http://%s:%d", registration.Address, registration.Port)
check.Timeout &＃61; "5s" //超时
check.Interval &＃61; "5s" //健康检查频率
check.DeregisterCriticalServiceAfter &＃61; "30s" // 故障检查失败30s后 consul自动将注册服务删除
registration.Check &＃61; check
// 注册服务到consul
err &＃61; client.Agent().ServiceRegister(registration)

func Handler(w http.ResponseWriter, r *http.Request)
w.Write([]byte("pong"))



启动程序&＃xff1a;

go run main.go
go run main.go -command_port 22202


consul结果如下&＃xff1a;

promethus结果如下&＃xff1a;

grafana添加数据源&＃xff1a;


选择add data source后选择promethus

填写promethus server的地址&＃xff0c;底部有个save and test

当出现下面这种情况的时候说明数据源可用&＃xff0c;不可用的话请检查consul, 检查promethus是否发现节点&＃xff0c;以及服务端的 http://xxxxxxx:xxxx/metrics的地址是否能访问且有数据,正常情况如下&＃xff1a;

选中dashboard, 添加一个空的panel

底部方框输入: myapp_processed_ops_total&＃xff0c; 这是我们在代码里面添加的指标名称

最终显示结果如下

到此整个流程完成&＃xff0c; 在代码里面动态注册服务到consul&＃xff0c;promethus server通过consul拿到所有注册组件的地址&＃xff0c;然后主动去拉取数据&＃xff0c;最后grafana通过进行展示。