prometheus服务发现

静态服务发现

Prometheus yml配置文件配置服务发现

[root@localhost prometheus-2.3.2.linux-amd64]# cat prometheus.yml 
# my global config
global:
  scrape_interval:     15s # Set the scrape interval to every 15 seconds. Default is every 1 minute.
  evaluation_interval: 15s # Evaluate rules every 15 seconds. The default is every 1 minute.
  # scrape_timeout is set to the global default (10s).

# Alertmanager configuration
alerting:
  alertmanagers:
  - static_configs:
    - targets:
      # - alertmanager:9093

# Load rules once and periodically evaluate them according to the global 'evaluation_interval'.
rule_files:
  # - "first_rules.yml"
  # - "second_rules.yml"

# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
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'.

    static_configs:
    - targets: ['localhost:9090']
  - job_name: 'test'
    static_configs:
    - targets:
      - 192.168.64.169:9100

scrape_confugs-抓取服务发现主配置
static_configs-静态服务发现配置，job_name-指的是在Prometheus可视化端筛选的标签，默认会添加一个Prometheus的job_name，用于监控自身服务是否正常，一个job是指多个类型功能的集合。

上述是job_name为test的参数指标，可以在promQL用下述代码进行筛选。

node_cpu_seconds_total{job="test"}

指标参数metrics是在配置中已经默认给予，可自己修改，targets对应着Prometheus可视化端的targets，可视化端记录了被监控的节点和state信息。

静态服务发现配置如下：

# metrics_path defaults to '/metrics'
  - job_name: 'test'
    static_configs:
    - targets:
      - 192.168.64.169:9100

DNS服务发现(不太熟悉)

prometheus 基于DNS的目标发现
DNS服务发现依赖于查询A、AAAA或SRV DNS记录。

1、基于 SRV 记录发现

scrape_configs:
  - job_name: 'webapp'
  dns_sd_configs:
    - names: [‘_prometheus._tcp.shhnwangjian.com]

注意：_prometheus 为服务名称，_tcp 为协议， shhnwangjian.com 为域名

2、基于 A 记录
需要编辑bind服务的named配置文件形成A记录

- job_name: 'webapp'
  dns_sd_configs:
    - names: [ 'ops.shhnwangjian.cn']
      type: A
      port: 9090

文件服务发现

准备工作

[root@localhost prometheus-2.3.2.linux-amd64]# mkdir target
[root@localhost prometheus-2.3.2.linux-amd64]# cd target/
[root@localhost target]# vim nodes-only.yaml
[root@localhost target]# cat nodes-only.yaml 
  - targets:
    - localhost:9090
    labels:
      app:prometheus
      job:prometheus
  - targets:
    - 192.168.64.169:9100
    labels:
      app:node-exporter
      job:test

 - job_name: node
   file_sd_configs
   - files:
     - targets/nodes-*.yaml
     refresh_interval: 2m  #每隔俩分钟加载一次

静态服务发现有一个很大的弊端，若你Prometheus.yml文件的targets进行修改，可视化端不会实时响应修改后的配置，使用文件服务发现，不需要依赖任何组件，refresh_interval可以解决静态服务发现不及时响应的弊端。

consul服务发现

让大量的node在consul进行注册，让Prometheus去读取注册信息
下载地址 https://www.consul.io/downloads/
安装 consul

unzip consul_1.9.2_linux_amd64.zip -d /usr/local/bin/

启动开发者模式 端口是8500

mkdir /consul/data -pv
consul agent -dev -ui -data-dir=/consul/data -config-dir=/etc/consul/ -client=0.0.0.0 

在/etc/sonsul 写对应的json文件

{
	"services":[
	{
	  "id":"node_exporter01",  
	  "name":"node01",  #在consul ui展示的名称
	  "address":"192.168.64.169",
	  "port":9100,
	  "tags":["nodes"],  #在Prometheus进行筛选
	  "checks":[{
	    "http":"http://192.168.64.146:9100/mestics",
	    "interval":"5s"
	  }]
	},
	{
	  "id":"node_exporter02",
      "name":"node02",
      "address":"192.168.64.169",
      "prot":9100,
      "tags":["nodes"],
      "checks":[{
        "http":"http://192.168.64.168:9100/mestics",
        "interval":"5s"
      }]
	}]
}

Prometheus.yml文件内 发现机制变成consul_sd_config
配置如下：

- job_name:"test"
  consul_sd_configs:
  - server: "consul的ui地址(ip:port)"
    tags:   #是/etc/consul/_.json内定义的tags和值
    - "nodes"
    refresh_interval: 2m #2m加载一下配置

再consul安装path内

consul reload

基于kubernetesAPI对各类资源包含Node，Service，Endpoint，Pod和Ingress等资源类型下相应的个资源对象是做target，持续监视相关资源变动，内容有点多，下个博客再介绍。