alertmanager实现自定义WebHook及路由(route)分组实战
作者:尹正杰
版权声明:原创作品,谢绝转载!否则将追究法律责任。
目录
一.alertmanager环境部署
1.alertmanager核心功能
deduplicating:
Prometheus产生同一条报警发送给多个alertmanager去重后再发送。
grouping:
告警可以分组处理,同一个组共享等待时长等参数,可以做告警聚合。
route:
路由匹配书,可以理解为告警订阅。
silencing:
灵活的告警静默,如按tag。
inhibition:
如果某些其它报警已经触发,则抑制某些警报的通知如机器down,上面的进程down告警不触发。
HA:
基于gossip可以实现alertmanager的高可用性。
2.部署alertmanager
1.下载alertmanager
[root@alertmanager44 ~]# wget https://github.com/prometheus/alertmanager/releases/download/v0.27.0/alertmanager-0.27.0.linux-amd64.tar.gz
2.解压软件包
[root@alertmanager44 ~]# mkdir -pv /yinzhengjie/{softwares,data}
3.解压软件包
[root@alertmanager44 ~]# tar xf alertmanager-0.27.0.linux-amd64.tar.gz -C /yinzhengjie/softwares/
4.编写启动脚本
[root@alertmanager44 ~]# cat > /lib/systemd/system/alertmanager.service <<EOF
[Unit]
Description=yinzhengjie's alertmanager Server
Documentation=https://www.cnblogs.com/yinzhengjie
After=network.target
[Service]
ExecStart=/yinzhengjie/softwares/alertmanager-0.27.0.linux-amd64/alertmanager --config.file=/yinzhengjie/softwares/alertmanager-0.27.0.linux-amd64/alertmanager.yml --storage.path=/yinzhengjie/data/alertmanager
[Install]
WantedBy=multi-user.target
EOF
4.启动服务
[root@alertmanager44 ~]# systemctl daemon-reload
[root@alertmanager44 ~]#
[root@alertmanager44 ~]# systemctl enable --now alertmanager.service
Created symlink /etc/systemd/system/multi-user.target.wants/alertmanager.service → /lib/systemd/system/alertmanager.service.
[root@alertmanager44 ~]#
5.访问alertmanager的WebUI
http://10.0.0.44:9093/
3.默认配置文件解读
# 定义的是全局配置
global:
# 如果一个告警不包括EndsAt,经过此时间后,如果尚未更新警报,则可以将警报声明为已恢复。
# 这对Prometheus的报警没用任何影响,因为它们始终包含EndsAt
resolve_timeout: 5m
# 配置默认的http,表示如果下面webhook为空的时候用此配置。
http_config:
follow_redirects: true
enable_http2: true
# SMTP配置。
smtp_hello: localhost
smtp_require_tls: true
# 几个默认支持地址,企业常用的是企业微信,钉钉告警,邮件告警等媒介。
pagerduty_url: https://events.pagerduty.com/v2/enqueue
opsgenie_api_url: https://api.opsgenie.com/
wechat_api_url: https://qyapi.weixin.qq.com/cgi-bin/
victorops_api_url: https://alert.victorops.com/integrations/generic/20131114/alert/
telegram_api_url: https://api.telegram.org
webex_api_url: https://webexapis.com/v1/messages
# 配置路由
route:
# 代表路由树的默认receiver,如果匹配不中路由就走默认的路由。
receiver: web.hook
# 定义分组的名称
group_by:
- alertname
continue: false
# 代表新的报警最小聚合时间,第一次来的时候最短间隔。
group_wait: 30s
# 代表同一个组里面告警聚合时间,同一个group_by里面不同tag的聚合时间。
group_interval: 5m
# 代表同一个报警(label完全相同)的最小发送间隔。
repeat_interval: 1h
# 配置抑制告警规则
inhibit_rules:
#
- source_match:
severity: critical
target_match:
severity: warning
equal:
- alertname
- dev
- instance
# 接收者配置
receivers:
- name: web.hook
webhook_configs:
- send_resolved: true
http_config:
follow_redirects: true
enable_http2: true
url: <secret>
url_file: ""
max_alerts: 0
# 定义文本模板
templates: []
二.编写go程序充当告警触发端和接收端
1 发送告警的接口
接口地址:
https://prometheus.io/docs/alerting/latest/clients/
依赖的公共库:
https://github.com/prometheus/common
2 发送消息到alertmanager
package main
import (
"bytes"
"encoding/json"
"io/ioutil"
"log"
"net/http"
"github.com/prometheus/common/model"
)
func AlertSend(alertmanagerAddress string) {
labels := model.LabelSet{}
labels["alertname"] = "杰哥讲运维,报警测试"
labels["group"] = "yinzhengjie"
labels["severity"] = "2"
labels["job"] = "yinzhengjie-k8s"
annotations := model.LabelSet{}
annotations["value"] = "2024"
// Alert结构体的数据和官网一致(https://prometheus.io/docs/alerting/latest/clients/)
alerts := make([]*model.Alert, 0)
alert := &model.Alert{
Labels: labels,
Annotations: annotations,
GeneratorURL: "https://www.yinzhengjie.com:9090",
}
alerts = append(alerts, alert)
jsonStr, _ := json.Marshal(alerts)
req, err := http.NewRequest("POST", alertmanagerAddress, bytes.NewBuffer(jsonStr))
req.Header.Set("Content-Type", "application/json")
client := &http.Client{}
resp, err := client.Do(req)
if err != nil {
log.Printf("http.post.requst.err|url:%v|err:%v", alertmanagerAddress, err)
return
}
defer resp.Body.Close()
log.Printf("response Status:%v", resp.Status)
log.Printf("response Headers:%v", resp.Header)
body, _ := ioutil.ReadAll(resp.Body)
log.Printf("response Body:%v", string(body))
}
func main() {
alertmanagerAddress := "http://10.0.0.44:9093/api/v2/alerts"
AlertSend(alertmanagerAddress)
}
3.接受告警段代码
3.1 修改alertmanager的配置文件
[root@alertmanager44 ~]# cat /yinzhengjie/softwares/alertmanager-0.27.0.linux-amd64/alertmanager.yml
route:
group_by: ['alertname']
group_wait: 30s
group_interval: 5m
repeat_interval: 1h
receiver: 'web.hook'
receivers:
- name: 'web.hook'
webhook_configs:
- url: 'http://10.0.0.41:8888/alert'
http_config: {}
max_alerts: 0
send_resolved: true
inhibit_rules:
- source_match:
severity: 'critical'
target_match:
severity: 'warning'
equal: ['alertname', 'dev', 'instance']
[root@alertmanager44 ~]#
3.2 源代码
package main
import (
"errors"
"fmt"
"log"
"net/http"
"flag"
"github.com/gin-gonic/gin"
"github.com/prometheus/alertmanager/notify/webhook"
)
func AlertReceiveFunc(c *gin.Context) {
var msg webhook.Message
if err := c.BindJSON(&msg); err != nil {
c.JSON(400, errors.New("invalid args"))
return
}
baseMsg := fmt.Sprintf("状态: %s|报警条数: %d", msg.Status, len(msg.Alerts))
log.Printf("alertReceive|baseMsg:%+v", baseMsg)
for i := 0; i < len(msg.Alerts); i++ {
alert := msg.Alerts[i]
log.Printf("detail|%d%d|alert:%+v", i+1, len(msg.Alerts), alert)
}
c.JSON(http.StatusOK, "ok")
}
func main() {
listenAddr := flag.String("addr", ":8888", "WebUI expose port")
flag.Parse()
r := gin.Default()
r.POST("/alert", AlertReceiveFunc)
r.Run(*listenAddr)
}
3.3 测试结果验证
如上图所示,我们可以成功看到日志的输出信息。
三.alertmanager分组功能
1.启动3个webhook测试终端
如上图所示,我们启动3个不同的终端测试,用于模拟3个不同的Webhook。
[root@node-exporter42 yinzhengjie]# ./alert-receive --addr :6666
[root@node-exporter42 yinzhengjie]# ./alert-receive --addr :7777
[root@node-exporter42 yinzhengjie]# ./alert-receive --addr :8888
2.alertmanager配置分组
分组说明:
- alertmanager可以根据设置的路由将告警可以分组处理,发送给对应的接收端;
- 我们可以定义三个接收组:
- sre_system
定义系统组,接收所有服务器机器告警。
- sre_dba
定义数据库组,接收数据库服务器告警。
- sre_bigdata
定义大数据组,接收大数据集群服务器告警。
alertmanager配置案例:
[root@prometheus-server31 ~]# cat /yinzhengjie/softwares/alertmanager-0.27.0.linux-amd64/alertmanager.yml
route:
group_by: ['alertname']
group_wait: 1s
group_interval: 3s
repeat_interval: 1m
# 默认发给"sre_system"组用户
receiver: 'sre_system'
continue: false
# 配置子路由
routes:
- receiver: 'sre_dba'
match_re:
job: yinzhengjie_mysqld_exporter
# 建议将continue的值设置为true,表示当前的条件是否匹配,都将继续向下匹配规则
# 这样做的目的是将消息发给最后的系统组(sre_system)
continue: true
- receiver: 'sre_bigdata'
match_re:
job: yinzhengjie_bigdata_exporter
continue: true
- receiver: 'sre_system'
match_re:
job: .*
continue: true
receivers:
- name: 'sre_system'
webhook_configs:
- url: 'http://10.0.0.42:6666/alert'
http_config: {}
max_alerts: 0
send_resolved: true
- name: 'sre_dba'
webhook_configs:
- url: 'http://10.0.0.42:7777/alert'
http_config: {}
max_alerts: 0
send_resolved: true
- name: 'sre_bigdata'
webhook_configs:
- url: 'http://10.0.0.42:8888/alert'
http_config: {}
max_alerts: 0
send_resolved: true
[root@prometheus-server31 ~]#
3.修改Prometheus的配置文件
[root@prometheus-server31 ~]# vim /yinzhengjie/softwares/prometheus-2.53.2.linux-amd64/prometheus.yml
...
global:
scrape_interval: 3s
...
alerting:
alertmanagers:
- static_configs:
- targets:
- 127.0.0.1:9093
...
rule_files:
- "/yinzhengjie/softwares/prometheus-2.53.2.linux-amd64/yinzhengjie_rules.yml"
...
scrape_configs:
...
- job_name: "yinzhengjie_mysqld_exporter"
static_configs:
- targets: ["10.0.0.42:9100","10.0.0.43:9100"]
- job_name: "yinzhengjie_bigdata_exporter"
static_configs:
- targets: ["10.0.0.31:9100","10.0.0.41:9100"]
4.定义规则
1.编写规则文件
[root@prometheus-server31 ~]# cat /yinzhengjie/softwares/prometheus-2.53.2.linux-amd64/yinzhengjie_rules.yml
groups:
- name: xixi
rules:
- alert: yinzhengjie_mysqld_exporter-alert
expr: node_boot_time_seconds{instance="10.0.0.42:9100", job="yinzhengjie_mysqld_exporter"} > 0
labels:
severity: critical
blog: "https://www.cnblogs.com/yinzhengjie"
annotations:
summary: DBA机器异常
- name: haha
rules:
- alert: yinzhengjie_bigdata_exporter-alert
expr: node_boot_time_seconds{instance="10.0.0.41:9100", job="yinzhengjie_bigdata_exporter"} > 0
labels:
severity: warning
auther: 尹正杰
annotations:
summary: 大数据集群机器异常
[root@prometheus-server31 ~]#
2.检查配置文件是否正确
[root@prometheus-server31 ~]# /yinzhengjie/softwares/prometheus-2.53.2.linux-amd64/promtool check config /yinzhengjie/softwares/prometheus-2.53.2.linux-amd64/prometheus.yml
Checking /yinzhengjie/softwares/prometheus-2.53.2.linux-amd64/prometheus.yml
SUCCESS: 1 rule files found
SUCCESS: /yinzhengjie/softwares/prometheus-2.53.2.linux-amd64/prometheus.yml is valid prometheus config file syntax
Checking /yinzhengjie/softwares/prometheus-2.53.2.linux-amd64/yinzhengjie_rules.yml
SUCCESS: 2 rules found
[root@prometheus-server31 ~]#
5.测试生效
如上图所示,我们可以直接测试重启一下Prometheus和alertmanager服务,观察webhook接收的数据分组情况。
[root@prometheus-server31 ~]# systemctl restart prometheus-server.service alertmanager.service
当你的才华还撑不起你的野心的时候,你就应该静下心来学习。当你的能力还驾驭不了你的目标的时候,你就应该沉下心来历练。问问自己,想要怎样的人生。
欢迎交流学习技术交流,个人微信: "JasonYin2020"(添加时请备注来源及意图备注)
作者: 尹正杰, 博客: https://www.cnblogs.com/yinzhengjie/p/18540986