Ansible部署prometheus

  1. ansible-playbook -i host_file  service_deploy.yaml  -e "tgz=prometheus-2.25.2.linux-amd64.tar.gz" -e "app=prometheus"
  4. 查看页面

image.png

prometheus配置文件解析

  1. # 全局配置段
  2. global:
  3.   # 采集间隔 
  4.   scrape_interval:     15s # Set the scrape interval to every 15 seconds. Default is every 1 minute.
  5.   # 计算报警和预聚合间隔
  6.   evaluation_interval: 15s # Evaluate rules every 15 seconds. The default is every 1 minute.
  7.   # 采集超时时间
  8.   scrape_timeout: 10s 
  9.   # 查询日志，包含各阶段耗时统计
  10.   query_log_file: /opt/logs/prometheus_query_log
  11.   # 全局标签组
  12.   # 通过本实例采集的数据都会叠加下面的标签
  13.   external_labels:
  14.     account: 'huawei-main'
  15.     region: 'beijng-01'
  18. # Alertmanager信息段
  19. alerting:
  20.   alertmanagers:
  21.   - scheme: http
  22.     static_configs:
  23.     - targets:
  24.       - "localhost:9093"
  26. # 告警、预聚合配置文件段
  27. rule_files:
  28.     - /etc/prometheus/rules/record.yml
  29.     - /etc/prometheus/rules/alert.yml
  31. # 采集配置段
  32. scrape_configs:
  33.   # The job name is added as a label `job=<job_name>` to any timeseries scraped from this config.
  34.   - job_name: 'prometheus'
  36.     # metrics_path defaults to '/metrics'
  37.     # scheme defaults to 'http'.
  39.     static_configs:
  40.     - targets: ['localhost:9090']
  42. # 远程查询段
  43. remote_read:
  44.   # prometheus 
  45.   - url: http://prometheus/v1/read
  46.     read_recent: true
  48.   # m3db 
  49.   - url: "http://m3coordinator-read:7201/api/v1/prom/remote/read"
  50.     read_recent: true
  52. # 远程写入段
  53. remote_write:
  54.   - url: "http://m3coordinator-write:7201/api/v1/prom/remote/write"
  55.     queue_config:
  56.       capacity: 10000
  57.       max_samples_per_send: 60000
  58.     write_relabel_configs:
  59.       - source_labels: [__name__]
  60.         separator: ;
  61.         # 标签key前缀匹配到的drop
  62.         regex: '(kubelet_|apiserver_|container_fs_).*'
  63.         replacement: $1
  64.         action: drop

所以prometheus实例可以用来做下列用


对应的配置段		 用途
采集配置段 做采集器，数据保存在本地
采集配置段 + 远程写入段 做采集器+传输器，数据保存在本地+远端存储
远程查询段 做查询器，查询远端存储数据
采集配置段 + 远程查询段 做采集器+查询器，查询本地数据+远端存储数据
采集配置段 + Alertmanager信息段 + 告警配置文件段 做采集器+告警触发器，查询本地数据生成报警发往Alertmanager
远程查询段 + Alertmanager信息段 + 告警配置文件段 做远程告警触发器，查询远端数据生成报警发往Alertmanager
远程查询段+远程写入段 + 预聚合配置文件段 做预聚合指标，生成的结果集指标写入远端存储 
  1. 准备prometheus配置文件，配置采集两个node_exporter
  1. global:
  2.   scrape_interval: 15s
  3.   scrape_timeout: 10s
  4.   evaluation_interval: 15s
  5. alerting:
  6.   alertmanagers:
  7.   - scheme: http
  8.     timeout: 10s
  9.     api_version: v1
  10.     static_configs:
  11.     - targets: []
  12. scrape_configs:
  13. - job_name: prometheus
  14.   honor_timestamps: true
  15.   scrape_interval: 15s
  16.   scrape_timeout: 10s
  17.   metrics_path: /metrics
  18.   scheme: http
  19.   static_configs:
  20.   - targets:
  21.     - localhost:9090
  22. - job_name: node_exporter
  23.   honor_timestamps: true
  24.   scrape_interval: 15s
  25.   scrape_timeout: 10s
  26.   metrics_path: /metrics
  27.   scheme: http
  28.   static_configs:
  29.   - targets:
  30.     - 172.16.58.79:9100
  31.     - 172.16.58.78:9100
  35. 热更新配置
  36. # 命令行开启  --web.enable-lifecycle
  37. curl -X POST http://localhost:9090/-/reload 
  39. 页面查看targets up情况

image.png
解说targets页面

  • job分组情况
  • endpoint实例地址
  • state采集是否成功
  • label标签组
  • Last Scrape 上次采集到现在的间隔时间
  • Scrape Duration 上次采集耗时
  • Error 采集错误

    通过api获取targets 详情

    ```yaml

    coding=UTF-8

import requests

def print_targets(targets): index = 1 all = len(targets) for i in targets: scrapeUrl = i.get(“scrapeUrl”) state = i.get(“health”) labels = i.get(“labels”) lastScrape = i.get(“lastScrape”) lastScrapeDuration = i.get(“lastScrapeDuration”) lastError = i.get(“lastError”) if state==”up”: up_type = “正常” else: up_type = “异常” msg = “状态:{} num:{}/{} endpoint:{} state:{} labels:{} lastScrape:{} lastScrapeDuration:{} lastError:{}”.format(

  1.         up_type,
  2.         index,
  3.         all,
  4.         scrapeUrl,
  5.         state,
  6.         str(labels),
  7.         lastScrape,
  8.         lastScrapeDuration,
  9.         lastError,
  11.     )
  12.     print(msg)
  13.     index+=1

def get_targets(t): f_data = {} try: uri = ‘http://{}/api/v1/targets'.format(t) res = requests.get(uri)

  1.     data = res.json().get("data")
  2.     activeTargets = data.get("activeTargets")
  3.     droppedTargets = data.get("droppedTargets")
  5.     ups = []
  6.     downs = []
  7.     print_targets(activeTargets)
  8.     print_targets(droppedTargets)
  11. except Exception as e:
  12.     print(e)

get_targets(“prometheus.master01.wiswrt.com:9090”)

``` image.png