Prometheus客户端docker监控cAdvisor

爱被打了一巴掌 2022-09-12 04:50 316阅读 0赞

从上一篇文章讲到客户端节点监控需要用node\_exporter，docker监控也需要用一个exporter，这个exporter就是cAdvisor，类似zabbix的各种agent  
cadvisor官网：[GitHub - google/cadvisor: Analyzes resource usage and performance characteristics of running containers.][GitHub - google_cadvisor_ Analyzes resource usage and performance characteristics of running containers.]

下面讲正题

1.下载安装

官网方案，直接用docker run运行

sudo docker run \
      --volume=/:/rootfs:ro \
      --volume=/var/run:/var/run:ro \
      --volume=/sys:/sys:ro \
      --volume=/var/lib/docker/:/var/lib/docker:ro \
      --volume=/dev/disk/:/dev/disk:ro \
      --publish=9080:8080 \
      --detach=true \
      --name=cadvisor \
      --privileged \
      --device=/dev/kmsg \
      gcr.io/cadvisor/cadvisor:latest

实际上gcr.io是谷歌的，谷歌这个域名是访问不下来的。所以需要一个折中方案

下载离线包：[cadvisor.tar-kubernetes文档类资源-CSDN下载][cadvisor.tar-kubernetes_-CSDN]

然后导入镜像

docker load -i cadvisor.tar

查看导入的target名和版本

docker image ls

可以看到导入的名为google/cadvisor:latest

然后执行docker run，把后面的gcr.io一串换成google/cadvisor:latest

然后执行docker ps能看到cadvisor这个进程

![watermark_type_ZHJvaWRzYW5zZmFsbGJhY2s_shadow_50_text_Q1NETiBAZ3NsczIwMDgwOA_size_20_color_FFFFFF_t_70_g_se_x_16][]

那个command全名是 "/usr/bin/cadvisor -logtostderr" 有些镜像是import导入的，需要指定这个command

之后我们打开

[http://10.24.11.202:9080/containers/][http_10.24.11.202_9080_containers]

ip换成自己的，可以看到容器状态

![watermark_type_ZHJvaWRzYW5zZmFsbGJhY2s_shadow_50_text_Q1NETiBAZ3NsczIwMDgwOA_size_20_color_FFFFFF_t_70_g_se_x_16 1][]

也可以用api看容器具体情况  
[http://10.24.11.202:9080/api/v1.0/containers/docker/988b9b94904649c6ce0bdb97e0a5f8b23c998f86435ef0074e653580c4d2f11f][http_10.24.11.202_9080_api_v1.0_containers_docker_988b9b94904649c6ce0bdb97e0a5f8b23c998f86435ef0074e653580c4d2f11f]

不过我们监控关注的是metrics

[http://10.24.11.202:9080/metrics][http_10.24.11.202_9080_metrics]

![watermark_type_ZHJvaWRzYW5zZmFsbGJhY2s_shadow_50_text_Q1NETiBAZ3NsczIwMDgwOA_size_20_color_FFFFFF_t_70_g_se_x_16 2][]

2.服务端prometheus.yml添加节点信息

编辑

vi /usr/local/prometheus/prometheus.yml

添加

- job_name: "202docker"
        static_configs:
          - targets: ["10.24.11.202:9080"]

名称自己起，targets需要配置客户端的ip端口号

重启服务端

systemctl restart prometheus

![watermark_type_ZHJvaWRzYW5zZmFsbGJhY2s_shadow_50_text_Q1NETiBAZ3NsczIwMDgwOA_size_20_color_FFFFFF_t_70_g_se_x_16 3][]

3.Grafana添加仪表

然后导入仪表盘插件，先从[官网][Link 1]找模板

这里推荐复制id 179 (Docker and Host Monitoring w/ Prometheus)

导入过程跟上一篇一样，仪表如图

![watermark_type_ZHJvaWRzYW5zZmFsbGJhY2s_shadow_50_text_Q1NETiBAZ3NsczIwMDgwOA_size_20_color_FFFFFF_t_70_g_se_x_16 4][]

[GitHub - google_cadvisor_ Analyzes resource usage and performance characteristics of running containers.]: https://github.com/google/cadvisor
[cadvisor.tar-kubernetes_-CSDN]: https://download.csdn.net/download/SiKongPop/12001034
[watermark_type_ZHJvaWRzYW5zZmFsbGJhY2s_shadow_50_text_Q1NETiBAZ3NsczIwMDgwOA_size_20_color_FFFFFF_t_70_g_se_x_16]: /images/20220828/41bbac6bad2f4255987867550cc436d2.png
[http_10.24.11.202_9080_containers]: http://10.24.11.202:9080/containers/
[watermark_type_ZHJvaWRzYW5zZmFsbGJhY2s_shadow_50_text_Q1NETiBAZ3NsczIwMDgwOA_size_20_color_FFFFFF_t_70_g_se_x_16 1]: /images/20220828/44e4ae89cb02416e90291ce669c2bba8.png
[http_10.24.11.202_9080_api_v1.0_containers_docker_988b9b94904649c6ce0bdb97e0a5f8b23c998f86435ef0074e653580c4d2f11f]: http://10.24.11.202:9080/api/v1.0/containers/docker/988b9b94904649c6ce0bdb97e0a5f8b23c998f86435ef0074e653580c4d2f11f
[http_10.24.11.202_9080_metrics]: http://10.24.11.202:9080/metrics
[watermark_type_ZHJvaWRzYW5zZmFsbGJhY2s_shadow_50_text_Q1NETiBAZ3NsczIwMDgwOA_size_20_color_FFFFFF_t_70_g_se_x_16 2]: /images/20220828/a6d8bc4603364244a0840de73ee1a693.png
[watermark_type_ZHJvaWRzYW5zZmFsbGJhY2s_shadow_50_text_Q1NETiBAZ3NsczIwMDgwOA_size_20_color_FFFFFF_t_70_g_se_x_16 3]: /images/20220828/dde140841bee40b9949fd498b347c851.png
[Link 1]: https://grafana.com/grafana/dashboards
[watermark_type_ZHJvaWRzYW5zZmFsbGJhY2s_shadow_50_text_Q1NETiBAZ3NsczIwMDgwOA_size_20_color_FFFFFF_t_70_g_se_x_16 4]: /images/20220828/d3e58857a572476d9271568e4b807dc4.png