码迷，mamicode.com
首页 > 其他好文 > 详细

Prometheus+Node_exporter+Grafana+Alertmanager 监控部署

时间：2019-11-12 01:02:42      阅读：69      评论：0      收藏：0      [点我收藏+]

标签：clone   全局配置   备份   实例   fir   wan   bytes   ora   使用率   

一、Prometheus 安装及配置

1、下载及解压安装包

cd /usr/local/src/

export VER="2.13.1"
wget https://github.com/prometheus/prometheus/releases/download/v${VER}/prometheus-${VER}.linux-amd64.tar.gz

mkdir -p /data0/prometheus 
groupadd prometheus
useradd -g prometheus prometheus -d /data0/prometheus

tar -xvf prometheus-${VER}.linux-amd64.tar.gz
cd /usr/local/src/
mv prometheus-${VER}.linux-amd64 /data0/prometheus/prometheus_server

cd /data0/prometheus/prometheus_server/
mkdir -p {data,config,logs,bin} 
mv prometheus promtool bin/
mv prometheus.yml config/

chown -R prometheus.prometheus /data0/prometheus

2 、设置环境变量

vim /etc/profile

PATH=/data0/prometheus/prometheus_server/bin:$PATH:$HOME/bin

source /etc/profile

3、检查配置文件

promtool check config /data0/prometheus/prometheus_server/config/prometheus.yml

Checking /data0/prometheus/prometheus_server/config/prometheus.yml
  SUCCESS: 0 rule files found

4、创建 prometheus.service 的 systemd unit 文件


#命令行测试机器人发送消息，验证是否可以发送成功，有的时候prometheus-webhook-dingtalk会报422的错误，就是因为钉钉的安全限制(这里的安全策略是发送消息，必须包含prometheus才可以正常发送)
curl -H "Content-Type: application/json" -d ‘{"msgtype":"text","text":{"content":"prometheus alert test"}}‘ https://oapi.dingtalk.com/robot/send?access_token=18f977769d50518e9d4f99a0d5dc1376f05615b61ea3639a87f106459f75b5c9

curl -H "Content-Type: application/json" -d ‘{"msgtype":"text","text":{"content":"prometheus alert test"}}‘ https://oapi.dingtalk.com/robot/send?access_token=11a0496d0af689d56a5861ae34dc47d9f1607aee6f342747442cc83e36715223

4.1、常规服务

sudo tee /etc/systemd/system/prometheus.service <<-‘EOF‘
[Unit]
Description=Prometheus
Documentation=https://prometheus.io/
After=network.target

[Service]
Type=simple
User=prometheus
ExecStart=/data0/prometheus/prometheus_server/bin/prometheus --config.file=/data0/prometheus/prometheus_server/config/prometheus.yml --storage.tsdb.path=/data0/prometheus/prometheus_server/data --storage.tsdb.retention=60d
Restart=on-failure

[Install]
WantedBy=multi-user.target
EOF

systemctl enable prometheus.service
systemctl stop prometheus.service
systemctl restart prometheus.service
systemctl status prometheus.service

4.2、使用 supervisor 管理 prometheus_server

yum install -y epel-release supervisor

sudo tee /etc/supervisord.d/prometheus.ini<<-"EOF"
[program:prometheus]
# 启动程序的命令;
command = /data0/prometheus/prometheus_server/bin/prometheus --config.file=/data0/prometheus/prometheus_server/config/prometheus.yml --storage.tsdb.path=/data0/prometheus/prometheus_server/data --storage.tsdb.retention=60d
# 在supervisord启动的时候也自动启动;
autostart = true
# 程序异常退出后自动重启;
autorestart = true
# 启动5秒后没有异常退出，就当作已经正常启动了;
startsecs = 5
# 启动失败自动重试次数，默认是3;
startretries = 3
# 启动程序的用户;
user = prometheus
# 把stderr重定向到stdout，默认false;
redirect_stderr = true
# 标准日志输出;
stdout_logfile=/data0/prometheus/prometheus_server/logs/out-prometheus.log
# 错误日志输出;
stderr_logfile=/data0/prometheus/prometheus_server/logs/err-prometheus.log
# 标准日志文件大小，默认50MB;
stdout_logfile_maxbytes = 20MB
# 标准日志文件备份数;
stdout_logfile_backups = 20
EOF

systemctl daemon-reload
systemctl enable supervisord
systemctl stop supervisord
systemctl restart supervisord
supervisorctl restart prometheus
supervisorctl status

5、prometheus.yml 配置文件

#创建Alertmanager告警规则文件
mkdir -p /data0/prometheus/prometheus_server/rules/
touch /data0/prometheus/prometheus_server/rules/node_down.yml
touch /data0/prometheus/prometheus_server/rules/memory_over.yml
touch /data0/prometheus/prometheus_server/rules/disk_over.yml
touch /data0/prometheus/prometheus_server/rules/cpu_over.yml

#prometheus配置文件
cat > /data0/prometheus/prometheus_server/config/prometheus.yml << \EOF
# my global config
global:
  scrape_interval: 15s # 设置抓取(pull)时间间隔，默认是1m
  evaluation_interval: 15s # 设置rules评估时间间隔，默认是1m
  # scrape_timeout is set to the global default (10s).

# 告警管理配置，默认配置
alerting:
  alertmanagers:
  - static_configs:
    - targets:
      - 192.168.56.11:9093 # 这里修改为 alertmanagers 的地址

# 加载rules，并根据设置的时间间隔定期评估
rule_files:
# - "first_rules.yml"
# - "second_rules.yml"
  - "/data0/prometheus/prometheus_server/rules/node_down.yml"                 # 实例存活报警规则文件
  - "/data0/prometheus/prometheus_server/rules/memory_over.yml"               # 内存报警规则文件
  - "/data0/prometheus/prometheus_server/rules/disk_over.yml"                 # 磁盘报警规则文件
  - "/data0/prometheus/prometheus_server/rules/cpu_over.yml"                  # cpu报警规则文件

# 抓取(pull)，即监控目标配置
# 默认只有主机本身的监控配置
scrape_configs:
  # The job name is added as a label `job=<job_name>` to any timeseries scraped from this config.
  - job_name: ‘prometheus‘

    # metrics_path defaults to ‘/metrics‘
    # scheme defaults to ‘http‘.

    # 可覆盖全局配置设置的抓取间隔，由15秒重写成5秒。
    scrape_interval: 10s

    static_configs:
      - targets: [‘localhost:9090‘, ‘localhost:9100‘]

  - job_name: ‘DMC_HOST‘
    file_sd_configs:
      - files: [‘./hosts.json‘]  
      # 被监控的主机，可以通过static_configs罗列所有机器，这里通过file_sd_configs参数加载文件的形式读取
      # 被监控的主机，可以json或yaml格式书写，我这里以json格式书写，target里面写监控机器的ip，labels非必须，可以由你自己定
EOF

#file_sd_configs参数形式配置主机列表
cat > /data0/prometheus/prometheus_server/config/hosts.json << \EOF
[
{
"targets": [
  "192.168.56.11:9100",
  "192.168.56.12:9100",
  "192.168.56.13:9100"
],
"labels": {
    "service": "db_node"
    }
},
{
"targets": [
  "192.168.56.14:9100",
  "192.168.56.15:9100",
  "192.168.56.16:9100"
],
"labels": {
    "service": "web_node"
    }
}
]
EOF

# 服务器存活报警
cat > /data0/prometheus/prometheus_server/rules/node_down.yml <<\EOF
groups:
- name: 实例存活告警规则
  rules:
  - alert: 实例存活告警
    expr: up == 0
    for: 1m
    labels:
      user: prometheus
      severity: warning
    annotations:
      description: "{{ $labels.instance }} of job {{ $labels.job }} has been down for more than 1 minutes."
EOF

# mem报警
cat > /data0/prometheus/prometheus_server/rules/memory_over.yml <<\EOF
groups:
- name: 内存报警规则
  rules:
  - alert: 内存使用率告警
    expr: (node_memory_MemTotal_bytes - (node_memory_MemFree_bytes+node_memory_Buffers_bytes+node_memory_Cached_bytes )) / node_memory_MemTotal_bytes * 100 > 80
    for: 1m
    labels:
      user: prometheus
      severity: warning
    annotations:
      description: "服务器: 内存使用超过80%！(当前值: {{ $value }}%)"
EOF

# disk报警
cat > /data0/prometheus/prometheus_server/rules/disk_over.yml <<\EOF
groups:
- name: 磁盘报警规则
  rules:
  - alert: 磁盘使用率告警
    expr: (node_filesystem_size_bytes - node_filesystem_avail_bytes) / node_filesystem_size_bytes * 100 > 80
    for: 1m
    labels:
      user: prometheus
      severity: warning
    annotations:
      description: "服务器: 磁盘设备: 使用超过80%！(挂载点: {{ $labels.mountpoint }} 当前值: {{ $value }}%)"
EOF

# cpu报警
cat > /data0/prometheus/prometheus_server/rules/cpu_over.yml <<\EOF
groups:
- name: CPU报警规则
  rules:
  - alert: CPU使用率告警
    expr: 100 - (avg by (instance)(irate(node_cpu_seconds_total{mode="idle"}[1m]) )) * 100 > 90
    for: 1m
    labels:
      user: prometheus
      severity: warning
    annotations:
      description: "服务器: CPU使用超过90%！(当前值: {{ $value }}%)"
EOF

6、查看 UI

Prometheus 自带有简单的 UI

http://192.168.56.11:9090/

http://192.168.56.11:9090/targets
http://192.168.56.11:9090/graph

二、node_exporter 安装及配置

1、下载及解压安装包

cd /usr/local/src/

export VER="0.18.1"
wget https://github.com/prometheus/node_exporter/releases/download/v${VER}/node_exporter-${VER}.linux-amd64.tar.gz

mkdir -p /data0/prometheus 
groupadd prometheus
useradd -g prometheus prometheus -d /data0/prometheus

tar -xvf node_exporter-${VER}.linux-amd64.tar.gz
cd /usr/local/src/
mv node_exporter-${VER}.linux-amd64 /data0/prometheus/node_exporter

chown -R prometheus.prometheus /data0/prometheus

2、创建 node_exporter.service的 systemd unit 文件

  • centos 下创建服务
cat > /usr/lib/systemd/system/node_exporter.service <<EOF
[Unit]
Description=node_exporter
Documentation=https://prometheus.io/
After=network.target

[Service]
Type=simple
User=prometheus
ExecStart=/data0/prometheus/node_exporter/node_exporter
Restart=on-failure

[Install]
WantedBy=multi-user.target
EOF
  • ubuntu 下创建服务
cat > /etc/systemd/system/node_exporter.service <<EOF
[Unit]
Description=node_exporter
Documentation=https://prometheus.io/
After=network.target

[Service]
Type=simple
User=prometheus
ExecStart=/data0/prometheus/node_exporter/node_exporter
Restart=on-failure

[Install]
WantedBy=multi-user.target
EOF

3、启动服务

systemctl daemon-reload
systemctl stop node_exporter.service
systemctl enable node_exporter.service
systemctl restart node_exporter.service

4、运行状态

systemctl status node_exporter.service

5、客户监控端数据汇报

访问：http://192.168.56.11:9100/metrics 查看从 exporter 具体能抓到的数据。

三、部署 Alertmanager 钉钉报警

1、下载及解压安装包

cd /usr/local/src/

export VER="0.19.0"
wget https://github.com/prometheus/alertmanager/releases/download/v${VER}/alertmanager-${VER}.linux-amd64.tar.gz

mkdir -p /data0/prometheus 
groupadd prometheus
useradd -g prometheus prometheus -d /data0/prometheus

tar -xvf alertmanager-${VER}.linux-amd64.tar.gz
cd /usr/local/src/
mv alertmanager-${VER}.linux-amd64 /data0/prometheus/alertmanager

chown -R prometheus.prometheus /data0/prometheus

2、配置 Alertmanager

alertmanager 的 webhook 集成了钉钉报警，钉钉机器人对文件格式有严格要求，所以必须通过特定的格式转换，才能发送给你钉钉的机器人。有人已经t贴心的为大家写了转换插件，那我们也就直接拿来用吧!

( https://github.com/timonwong/prometheus-webhook-dingtalk.git )

cat >/data0/prometheus/alertmanager/alertmanager.yml<<-"EOF"
# 全局配置项
global:
  resolve_timeout: 5m # 处理超时时间，默认为5min

# 定义路由树信息
route:
  group_by: [alertname]  # 报警分组依据
  receiver: ops_notify   # 设置默认接收人
  group_wait: 30s        # 最初即第一次等待多久时间发送一组警报的通知
  group_interval: 60s    # 在发送新警报前的等待时间
  repeat_interval: 1h    # 重复发送告警时间。默认1h
  routes:

  - receiver: ops_notify  # 基础告警通知
    group_wait: 10s
    match_re:
      alertname: 实例存活告警|磁盘使用率告警   # 匹配告警规则中的名称发送

  - receiver: info_notify  # 消息告警通知
    group_wait: 10s
    match_re:
      alertname: 内存使用率告警|CPU使用率告警

# 定义基础告警接收者
receivers:
- name: ops_notify
  webhook_configs:
  - url: http://localhost:8060/dingtalk/ops_dingding/send 
    send_resolved: true  # 警报被解决之后是否通知

# 定义消息告警接收者
- name: info_notify
  webhook_configs:
  - url: http://localhost:8060/dingtalk/info_dingding/send 
    send_resolved: true

# 一个inhibition规则是在与另一组匹配器匹配的警报存在的条件下，使匹配一组匹配器的警报失效的规则。两个警报必须具有一组相同的标签。
inhibit_rules: 
  - source_match: 
      severity: ‘critical‘ 
    target_match: 
      severity: ‘warning‘ 
    equal: [‘alertname‘, ‘dev‘, ‘instance‘]
EOF

3、启动 Alertmanager

cat >/lib/systemd/system/alertmanager.service<<\EOF
[Unit]
Description=Prometheus: the alerting system
Documentation=http://prometheus.io/docs/
After=prometheus.service

[Service]
ExecStart=/data0/prometheus/alertmanager/alertmanager --config.file=/data0/prometheus/alertmanager/alertmanager.yml
Restart=always
StartLimitInterval=0
RestartSec=10

[Install]
WantedBy=multi-user.target
EOF

systemctl enable alertmanager.service
systemctl stop alertmanager.service
systemctl restart alertmanager.service
systemctl status alertmanager.service

#查看端口
netstat -anpt | grep 9093

4、将钉钉接入 Prometheus AlertManager WebHook

#命令行测试机器人发送消息，验证是否可以发送成功，有的时候prometheus-webhook-dingtalk会报422的错误，就是因为钉钉的安全限制(这里的安全策略是发送消息，必须包含prometheus才可以正常发送)
curl -H "Content-Type: application/json" -d ‘{"msgtype":"text","text":{"content":"prometheus alert test"}}‘ https://oapi.dingtalk.com/robot/send?access_token=18f977769d50518e9d4f99a0d5dc1376f05615b61ea3639a87f106459f75b5c9

curl -H "Content-Type: application/json" -d ‘{"msgtype":"text","text":{"content":"prometheus alert test"}}‘ https://oapi.dingtalk.com/robot/send?access_token=11a0496d0af689d56a5861ae34dc47d9f1607aee6f342747442cc83e36715223

4.1、二进制包方式部署插件

cd /usr/local/src/
export VER="0.3.0"
wget https://github.com/timonwong/prometheus-webhook-dingtalk/releases/download/v${VER}/prometheus-webhook-dingtalk-${VER}.linux-amd64.tar.gz
tar -zxvf prometheus-webhook-dingtalk-${VER}.linux-amd64.tar.gz
mv prometheus-webhook-dingtalk-${VER}.linux-amd64 /data0/prometheus/alertmanager/prometheus-webhook-dingtalk

#使用方法：prometheus-webhook-dingtalk --ding.profile=钉钉接收群组的值=webhook的值

cat > /etc/systemd/system/prometheus-webhook-dingtalk.service<<\EOF
[Unit]
Description=prometheus-webhook-dingtalk
After=network-online.target

[Service]
Restart=on-failure
ExecStart=/data0/prometheus/alertmanager/prometheus-webhook-dingtalk/prometheus-webhook-dingtalk           --ding.profile=ops_dingding=https://oapi.dingtalk.com/robot/send?access_token=18f977769d50518e9d4f99a0d5dc1376f05615b61ea3639a87f106459f75b5c9           --ding.profile=info_dingding=https://oapi.dingtalk.com/robot/send?access_token=11a0496d0af689d56a5861ae34dc47d9f1607aee6f342747442cc83e36715223          

[Install]
WantedBy=multi-user.target
EOF

systemctl daemon-reload
systemctl stop prometheus-webhook-dingtalk
systemctl restart prometheus-webhook-dingtalk
systemctl status prometheus-webhook-dingtalk

netstat -nltup|grep 8060

4.2、docker 方式部署插件

docker pull timonwong/prometheus-webhook-dingtalk:v0.3.0

#docker run -d --restart always -p 8060:8060 timonwong/prometheus-webhook-dingtalk:v0.3.0 --ding.profile="<web-hook-name>=<dingtalk-webhook>"

docker run -d --restart always -p 8060:8060 timonwong/prometheus-webhook-dingtalk:v0.3.0 --ding.profile="ops_dingding=https://oapi.dingtalk.com/robot/send?access_token=18f977769d50518e9d4f99a0d5dc1376f05615b61ea3639a87f106459f75b5c9" --ding.profile="info_dingding=https://oapi.dingtalk.com/robot/send?access_token=11a0496d0af689d56a5861ae34dc47d9f1607aee6f342747442cc83e36715223"

这里解释一下两个变量：

<web-hook-name> ：prometheus-webhook-dingtalk 支持多个钉钉 webhook，不同 webhook 就是靠名字对应到 URL 来做映射的。要支持多个钉钉 webhook，可以用多个 --ding.profile 参数的方式支持，例如：sudo docker run -d --restart always -p 8060:8060 timonwong/prometheus-webhook-dingtalk:v0.3.0 --ding.profile="webhook1=https://oapi.dingtalk.com/robot/send?access_token=token1" --ding.profile="webhook2=https://oapi.dingtalk.com/robot/send?access_token=token2"。而名字和 URL 的对应规则如下，ding.profile="webhook1=......"，对应的 API URL 为：http://localhost:8060/dingtalk/webhook1/send

<dingtalk-webhook>：这个就是之前获取的钉钉 webhook

4.3、源码方式部署插件

#安装golang环境
cd /usr/local/src/
wget https://dl.google.com/go/go1.13.4.linux-amd64.tar.gz
tar -zxvf go1.13.4.linux-amd64.tar.gz
mv go/ /usr/local/

#vim /etc/profile
export GOROOT=/usr/local/go
export PATH=$PATH:$GOROOT/bin

#添加环境变量GOPATH
mkdir -p /opt/path
export GOPATH=/opt/path

#若 $GOPATH/bin 没有加入$PATH中，你需要执行将其可执行文件移动到$GOBIN下
export GOPATH=/opt/path
export PATH=$PATH:$GOROOT/bin:$GOPATH/bin
source /etc/profile

#下载插件
cd /usr/local/src/
git clone https://github.com/timonwong/prometheus-webhook-dingtalk.git
cd prometheus-webhook-dingtalk
go get github.com/timonwong/prometheus-webhook-dingtalk/cmd/prometheus-webhook-dingtalk
make   #(make成功后，会产生一个prometheus-webhook-dingtalk二进制文件)

#将钉钉告警插件拷贝到alertmanager目录
cp prometheus-webhook-dingtalk /data0/prometheus/alertmanager/

#启动服务
nohup /data0/prometheus/alertmanager/prometheus-webhook-dingtalk/prometheus-webhook-dingtalk --ding.profile="ops_dingding=https://oapi.dingtalk.com/robot/send?access_token=18f977769d50518e9d4f99a0d5dc1376f05615b61ea3639a87f106459f75b5c9" --ding.profile="info_dingding=https://oapi.dingtalk.com/robot/send?access_token=11a0496d0af689d56a5861ae34dc47d9f1607aee6f342747442cc83e36715223" 2>&1 1>/tmp/dingding.log &

#检查端口
netstat -anpt | grep 8060

四、Grafana 安装及配置

1、下载及安装

cd /usr/local/src/

export VER="6.4.3"
wget https://dl.grafana.com/oss/release/grafana-${VER}-1.x86_64.rpm
yum localinstall -y grafana-${VER}-1.x86_64.rpm

2、启动服务

systemctl daemon-reload
systemctl enable grafana-server.service
systemctl stop grafana-server.service
systemctl restart grafana-server.service

3、访问 WEB 界面

默认账号/密码：admin/admin http://192.168.56.11:3000

4、Grafana 添加数据源

在登陆首页，点击"Configuration-Data Sources"按钮，跳转到添加数据源页面，配置如下：
Name: prometheus
Type: prometheus
URL: http://192.168.56.11:9090
Access: Server
取消Default的勾选，其余默认，点击"Add"，如下：

需要安装饼图的插件
grafana-cli plugins install grafana-piechart-panel
systemctl restart grafana-server.service

请确保安装后能正常添加饼图。

安装consul数据源插件
grafana-cli plugins install sbueringer-consul-datasource
systemctl restart grafana-server.service

五、替换 grafana 的 dashboards

https://grafana.com/grafana/dashboards/11074 基础监控-new

https://grafana.com/dashboards/8919 基础监控

https://grafana.com/dashboards/7362 数据库监控

参考文档：

https://www.jianshu.com/p/e59cfd15612e Centos 7 部署 Prometheus、Alertmanager、Grafana 监控 Linux 主机

https://juejin.im/entry/5c2c4a7f6fb9a049b82a90ee 使用 Prometheus 监控 Ceph

https://blog.csdn.net/xiegh2014/article/details/84936174 CentOS7.5 Prometheus2.5+Grafana5.4监控部署

https://www.cnblogs.com/smallSevens/p/7805842.html Grafana+Prometheus打造全方位立体监控系统

https://www.cnblogs.com/sfnz/p/6566951.html安装prometheus+grafana 监控 mysql redis kubernetes 等

https://blog.csdn.net/hzs33/article/details/86553259 prometheus+grafana 监控 mysql、canal服务器

作者：Lancger

github 地址：https://github.com/Lancger/opslinux/blob/master/monitor/Prometheus/1.Prometheus%2BNode_exporter%2BGrafana%2BAlertmanager%E7%9B%91%E6%8E%A7%E4%B8%80%E4%BD%93%E5%8C%96.md

Prometheus+Node_exporter+Grafana+Alertmanager 监控部署

标签：clone   全局配置   备份   实例   fir   wan   bytes   ora   使用率   

原文地址：https://blog.51cto.com/51reboot/2449530
踩
(0)
赞
(0)
   
举报
评论 一句话评论（0）
登录后才能评论！
分享档案
更多>
2021年07月29日 (22)
2021年07月28日 (40)
2021年07月27日 (32)
2021年07月26日 (79)
2021年07月23日 (29)
2021年07月22日 (30)
2021年07月21日 (42)
2021年07月20日 (16)
2021年07月19日 (90)
2021年07月16日 (35)
周排行
mamicode.com排行更多图片
更多
友情链接
兰亭集智   国之画   百度统计   站长统计   阿里云   chrome插件   新版天听网
关于我们 - 联系我们 - 留言反馈
© 2014 mamicode.com 版权所有  联系我们:gaon5@hotmail.com
迷上了代码！