【云原生 | Kubernetes 系列】—Prometheus 联邦

序号	服务器IP	角色
1	192.168.31.201	prometheus-server
2	192.168.31.121	prometheus-node1
3	192.168.31.122	prometheus-node2
4	192.168.31.123	prometheus-node3
5	192.168.31.101	k8s-node-exporter1
6	192.168.31.102	k8s-node-exporter2
7	192.168.31.103	k8s-node-exporter3
8	192.168.31.111	k8s-node-exporter4
9	192.168.31.112	k8s-node-exporter5
10	192.168.31.113	k8s-node-exporter6
11	192.168.31.114	k8s-node-exporter7

1. Prometheus搭建

1. 将安装包和service文件复制到3台联邦节点上

# scp prometheus-2.38.0.linux-amd64.tar.gz 192.168.31.121:/apps/
# scp /etc/systemd/system/prometheus.service 192.168.31.121:/etc/systemd/system/prometheus.service

# scp prometheus-2.38.0.linux-amd64.tar.gz 192.168.31.122:/apps/
# scp /etc/systemd/system/prometheus.service 192.168.31.122:/etc/systemd/system/prometheus.service

# scp prometheus-2.38.0.linux-amd64.tar.gz 192.168.31.123:/apps/
# scp /etc/systemd/system/prometheus.service 192.168.31.123:/etc/systemd/system/prometheus.service

2. 在3台联邦服务器上安装prometheus

root@zookeeper-3:/apps# cd /apps
root@zookeeper-3:/apps# tar xf prometheus-2.38.0.linux-amd64.tar.gz
root@zookeeper-3:/apps# ln -sf /apps/prometheus-2.38.0.linux-amd64 /apps/prometheus

2. 所有节点安装node-exporter

这些节点之前都装过了,就不再复述了
只要确认可以通过nose:9100/metrics访问到数据即可

3. 配置联邦

3.1 prometheus-node1

/apps/prometheus/prometheus.yml追加node-exporter配置

  - job_name: "node-exporter"
    static_configs:
      - targets: ["192.168.31.101:9100","192.168.31.102:9100","192.168.31.103:9100"]

启动服务后确认9090监听是否正常

root@zookeeper-1:/apps/prometheus# systemctl enable --now prometheus.service 
Created symlink /etc/systemd/system/multi-user.target.wants/prometheus.service → /etc/systemd/system/prometheus.service.
root@zookeeper-1:/apps/prometheus# ss -ntl|grep 9090
LISTEN  0       4096                    *:9090                 *:*      

在web页面上查看是否正常获取到数据

3.2 其他prometheus-node节点

prometheus-node2

  - job_name: "node-exporter"
    static_configs:
      - targets: ["192.168.31.111:9100","192.168.31.112:9100"]

prometheus-node3

  - job_name: "node-exporter"
    static_configs:
      - targets: ["192.168.31.113:9100","192.168.31.114:9100"]

4. 配置prometheus-server

修改配置prometheus.yml将以下配置加入scrape_configs:

scrape_configs:
  - job_name: "prometheus-federate"
    scrape_interval: 10s
    honor_labels: true
    metrics_path: '/federate'
    params:
      'match[]':
      - '{job="prometheus"}'
      - '{__name__=~"job:.*"}'
      - '{__name__=~"node.*"}'
    static_configs:
    - targets:
      - '192.168.31.121:9090'
      - '192.168.31.122:9090'
      - '192.168.31.123:9090'

重启prometheus服务后
在Targets下看到了联邦的3个node节点.

此时其他配置都已经删除了,所以只有这么一个配置.
查询node_load1时可以看到3个node分别收集的node信息在server上都能正常获取到

5. Grafana导入

8919