grafana对报警的支持真的很弱，而Prometheus提供的报警系统就强大很多

Prometheus将数据采集和报警分成了两个模块。报警规则配置在Prometheus Servers上，然后发送报警信息到AlertManger，然后我们的AlertManager就来管理这些报警信息，包括silencing、inhibition，聚合报警信息过后通过email、PagerDuty、HipChat、Slack 等方式发送消息提示。

让AlertManager提供服务总的来说就下面3步：

1.安装和配置AlertManger

2.配置Prometheus来和AlertManager通信

3.在Prometheus中创建报警规则

一个报警信息在生命周期内有下面3种状态：

1.inactive: 表示当前报警信息既不是firing状态也不是pending状态

2.pending: 表示在设置的阈值时间范围内被激活了

3.firing: 表示超过设置的阈值时间被激活了

Alertmanager配置文件

global:
  resolve_timeout: 5m
  # smtp配置
  smtp_from: "prom-alert@example.com"
  smtp_smarthost: 'email-smtp.us-west-2.amazonaws.com:465'
  smtp_auth_username: "user"
  smtp_auth_password: "pass"
  smtp_require_tls: true
templates:
  - '/data/alertmanager/templates/*.tmpl'
route:
  receiver: test1
  group_wait: 30s
  group_interval: 5m
  repeat_interval: 4h
  group_by: [alertname]
  routes:
# ads webhook
  - receiver: test1
    group_wait: 10s
    match:
      team: ads
# ops webhook
  - receiver: test2
    group_wait: 10s
    match:
      team: operations
receivers:
- name: test1
  email_configs:
  - to: '9935226@qq.com'
    headers: { Subject: "[ads] 报警邮件"} # 接收邮件的标题
  webhook_configs:
  - url: http://localhost:8060/dingtalk/ads/send

- name: test2
  email_configs:
  - to: '9935226@qq.com,deniss.wang@gmail.com'
    send_resolved: true
    headers: { Subject: "[ops] 报警邮件"} # 接收邮件的标题
  webhook_configs:
  - url: http://localhost:8060/dingtalk/ops/send
    # wx config
  wechat_configs:
  - corp_id: 'wwxxxxxxxxxxxxxx'
    api_url: 'https://qyapi.weixin.qq.com/cgi-bin/'
    send_resolved: true
    to_party: '2'
    agent_id: '1000002'
    api_secret: '1FvHxuGbbG35FYsuW0YyI4czWY/.2'

将Dingtalk接入 Prometheus AlertManager WebHook

在钉钉中申请钉钉机器人：

格式为：https://oapi.dingtalk.com/robot/send?access_token=11a0496d0af689d56a5861ae34dc47d9f1607aee6f342747123

二进制方式安装Dingtalk-Webhook插件插件

cd /usr/local/src/
wget https://github.com/timonwong/prometheus-webhook-dingtalk/releases/download/v0.3.0/prometheus-webhook-dingtalk-0.3.0.linux-amd64.tar.gz
tar -zxvf prometheus-webhook-dingtalk-0.3.0.linux-amd64.tar.gz
mv prometheus-webhook-dingtalk-0.3.0.linux-amd64 /data/alertmanager/webhook-dingtalk

# 创建Systemd webhook-dingtalk 服务
cat > /etc/systemd/system/webhook-dingtalk.service << EOF
[Unit]
Description=webhook-dingding
Documentation=https://prometheus.io/
After=network.target

[Service]
Type=simple
User=prometheus
ExecStart=/data/alertmanager/webhook-dingtalk/prometheus-webhook-dingtalk \
--ding.profile="ads=https://oapi.dingtalk.com/robot/send?access_token=284de68124e97420a2ee8ae1b8f12fabe3213213213" \
--ding.profile="ops=https://oapi.dingtalk.com/robot/send?access_token=8bce3bd11f7040d57d44caa5b6ef9417eab24e1123123123213"
Restart=on-failure

[Install]
WantedBy=multi-user.target
EOF

# 启动服务
systemctl enable webhook-dingtalk
systemctl start webhook-dingtalk
systemctl status webhook-dingtalk

# 查看端口是否正常
netstat -anplt|grep 8060
tcp6       0      0 :::8060                 :::*                    LISTEN      1635/prometheus-web