Prometheus简单部署

Prometheus主要两个组件，prometheus server和alertmanager

一、准备工作

本文的prometheus的版本为2.25.2

./prometheus —version

prometheus, version 2.25.2 (branch: HEAD, revision: bda05a23ada314a0b9806a362da39b7a1a4e04c3) build user: root@de38ec01ef10 build date: 20210316-18:07:52 go version: go1.15.10 platform: linux/amd64

 安装包github自行下载即可

二、Prometheus server 部署

1.prometheus server的supervisor配置文件

[program:prometheus]
command=/data/server/prometheus/prometheus --config.file=/data/server/prometheus/prometheus.yml --storage.tsdb.path=/data/prometheus_data --web.listen-address="0.0.0.0:9090" --storage.tsdb.min-block-duration=2h --storage.tsdb.max-block-duration=2h --storage.tsdb.retention.time=7d --web.enable-lifecycle
user=root
stderr_logfile = /var/log/supervisor/prometheus_err.log
stdout_logfile = /var/log/supervisor/prometheus_stdout.log
directory = /data/server/prometheus
autostart=true
autorestart=true
startsecs=5

2.prometheus server配置文件

vim prometheus.yml

global:
  external_labels:             #打一个额外的label
    scrapeReplica: 2
  scrape_interval:     1s # Set the scrape interval to every 15 seconds. Default is every 1 minute.
  evaluation_interval: 1s # Evaluate rules every 15 seconds. The default is every 1 minute.
alerting:
  alertmanagers:
  - static_configs:
    - targets: ['localhost:9093']      #填写alertmanager的IP和端口
rule_files:
  - "/data/server/prometheus-2.25.2.linux-amd64/rules/alert.yml"   #填写报警规则，触发后发给alertmanager
scrape_configs:
  - job_name: 'prometheus'
    static_configs:
    - targets: ['localhost:9090']        #静态配置，此处为收集本机prometheus的指标
  - job_name: 'web'
    consul_sd_configs:                   #基于consul的服务发现
    - server: '192.168.13.29:8500'       #consul节点的机器和端口
      services: ["web"]                  #需要监听的consul的哪个services
  - job_name: 'NodeExporter'
    static_configs:
    - targets: ['localhost:9100']

3.rule_files文件配置

此处示例一个简单的内存报警配置规则

groups:
- name: node_health
  rules:
  - alert: HighMemoryUsage           #报警名字
    expr: node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes < 0.9   #什么情况下会报警
    for: 1m                          #持续一分钟后才会触发报警
    labels:
      severity: warning              #自定义报警级别
    annotations:
      summary: High memory usage     #自己写报警信息

出发报警后的发送的邮件信息

三、alertmanager报警

1.alertmanager以supervisor启动

[program:alertmanager]
command=/data/server/alertmanager/alertmanager --config.file=/data/server/alertmanager/alertmanager.yml
user=root
stderr_logfile = /var/log/supervisor/alertmanager_err.log
stdout_logfile = /var/log/supervisor/alertmanager_stdout.log
directory = /data/server/alertmanager
autostart=true
autorestart=true
startsecs=5

2.alertmanager配置文件

vim alertmanager.yml

global:
  resolve_timeout: 5m      #在配置的超时时间内没有收到上次告警就会发出恢复邮件，恢复也是按分组发出; 分组会包涵恢复和告警信息
  smtp_smarthost: 'smtp.feishu.cn:465'
  smtp_from: 'zltest@xiaoniangao.com'        #用这个邮箱去发送报警信息
  smtp_auth_username: 'zltest@xiaoniangao.com'
  smtp_auth_password: 'xxxxxxxxxxxx'
  smtp_require_tls: false
route:
  group_by: ['alertname', 'cluster', 'service']   #满足group_by中定义的标签，那么这些报警将会合并为一个报警通知发送给receiver
  group_wait: 10s      #为了一次收集更多的信息，如果在等待的时间内收到了新的报警，这些报警会合并成一个通知发送给receiver
  group_interval: 10s  #定义相同group之间发送告警的时间间隔
  repeat_interval: 5m  #报警周期，发送一次报警后，下次5分钟后再发送一次报警
  receiver: 'mail'     #报警发送使用mail这个规则
receivers:
- name: 'mail'
  email_configs:
  - to: 'qinjunhang@xiaoniangao.com'     #报警接收人邮箱

至此prometheus的组件已经部署完成，但是怎么去收集监控数据呢，此时需要一个exporter去收集机器上的信息（就类比zabbix上的一个脚本，去采集机器指标），下面来部署一个主机监控的exporter

四、NodeExporter部署

1.NodeExporter以supervisor启动

[program:node-exporter]
command=/data/server/node_exporter/node_exporter
  --web.disable-exporter-metrics
  --collector.processes
  --collector.filesystem.ignored-mount-points='^/(dev|run|proc|sys|var/lib/docker/.+)($|/)'
stderr_logfile = /var/log/supervisor/node-exporter_err.log
stdout_logfile = /var/log/supervisor/node-exporter_stdout.log
directory = /data/server/node_exporter
user=root
autostart=true
autorestart=true
startsecs=5