prometheus 集羣

思路一

統一區域的監控目標，prometheus server兩臺監控相同的目標羣體。

改變後

上面這個變化對於監控目標端，會多出一倍的查詢請求，但在一臺prometheus server宕機的狀況下，能夠不影響監控。

思路二

這是一個金字塔式的層次結構，而不是分佈式層次結構。Prometheus 的抓取請求也會加載到prometheus work節點上，這是須要考慮的。

上面這種模式，準備3臺prometheus server進行搭建，這種方式work節點一臺宕機後，其它wokr節點不會去接手故障work節點的機器。

一、環境準備

192.168.31.151（primary）

192.168.31.144 （worker）

192.168.31.82（worker）

二、部署prometheus

cd /usr/loacl
tar -xvf prometheus-2.8.0.linux-amd64.tar.gz 
ln -s /usr/local/prometheus-2.8.0.linux-amd64 /usr/local/prometheus
cd /usr/local/prometheus;mkdir bin conf data
mv ./promtool bin
mv ./prometheus bin
mv ./prometheus.yml conf

三、worker節點配置（192.168.31.144）

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.
  external_labels:
    worker: 0

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

# Load rules once and periodically evaluate them according to the global 'evaluation_interval'.
rule_files:
  - "rules/*_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'
    static_configs:
    - targets:
      - 192.168.31.151:9090
      - 192.168.31.144:9090
      - 192.168.31.82:9090
    relabel_configs:
    - source_labels: [__address__]
      modulus: 2
      target_label: __tmp_hash
      action: hashmod
    - source_labels: [__tmp_hash]
      regex: ^0$
      action: keep
  - job_name: 'node_exporter'
    file_sd_configs:
      - files:
        - targets/nodes/*.json
        refresh_interval: 1m
    relabel_configs:
    - source_labels: [__address__]
      modulus: 2
      target_label: __tmp_hash
      action: hashmod
    - source_labels: [__tmp_hash]
      regex: ^0$
      action: keep
  - job_name: 'docker'
    file_sd_configs:
      - files:
        - targets/docker/*.json
        refresh_interval: 1m
    relabel_configs:
    - source_labels: [__address__]
      modulus: 2
      target_label: __tmp_hash
      action: hashmod
    - source_labels: [__tmp_hash]
      regex: ^0$
      action: keep
  - job_name: 'alertmanager'
    static_configs:
    - targets:
      - 192.168.31.151:9093
      - 192.168.31.144:9093
    relabel_configs:
    - source_labels: [__address__]
      modulus: 2
      target_label: __tmp_hash
      action: hashmod
    - source_labels: [__tmp_hash]
      regex: ^0$
      action: keep

worker節點配置（192.168.31.82）

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.
  external_labels:
    worker: 1

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

# Load rules once and periodically evaluate them according to the global 'evaluation_interval'.
rule_files:
  - "rules/*_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'
    static_configs:
    - targets:
      - 192.168.31.151:9090
      - 192.168.31.144:9090
      - 192.168.31.82:9090
    relabel_configs:
    - source_labels: [__address__]
      modulus: 2
      target_label: __tmp_hash
      action: hashmod
    - source_labels: [__tmp_hash]
      regex: ^1$
      action: keep
  - job_name: 'node_exporter'
    file_sd_configs:
      - files:
        - targets/nodes/*.json
        refresh_interval: 1m
    relabel_configs:
    - source_labels: [__address__]
      modulus: 2
      target_label: __tmp_hash
      action: hashmod
    - source_labels: [__tmp_hash]
      regex: ^1$
      action: keep
  - job_name: 'docker'
    file_sd_configs:
      - files:
        - targets/docker/*.json
        refresh_interval: 1m
    relabel_configs:
    - source_labels: [__address__]
      modulus: 2
      target_label: __tmp_hash
      action: hashmod
    - source_labels: [__tmp_hash]
      regex: ^1$
      action: keep
  - job_name: 'alertmanager'
    static_configs:
    - targets:
      - 192.168.31.151:9093
      - 192.168.31.144:9093
    relabel_configs:
    - source_labels: [__address__]
      modulus: 2
      target_label: __tmp_hash
      action: hashmod
    - source_labels: [__tmp_hash]
      regex: ^1$
      action: keep

primary節點配置（192.168.31.151）

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:
      - 192.168.31.151:9093
      - 192.168.31.144:9093

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

# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
scrape_configs:
  - job_name: 'node_workers'
    file_sd_configs:
      - files:
        - 'targets/workers/*.json'
        refresh_interval: 5m
    honor_labels: true
    metrics_path: /federate
    params:
      'match[]':
         - '{__name__=~"^instance:.*"}'

cat ./targets/workers/workers.json 

[{
  "targets": [
    "192.168.31.144:9090",
    "192.168.31.82:9090"
  ]
}]