Prometheus + Grafana 快速上手

Prometheus[1] + Grafana[2] 快速上手，监控主机的 CPU, GPU, MEM, IO 等状态。

前提

• Docker[3]

客户端

Node Exporter

用于采集 UNIX 内核主机的数据，这里下载[4]并解压：

wget https://github.com/prometheus/node_exporter/releases/download/v1.1.2/node_exporter-1.1.2.linux-amd64.tar.gz
tar xvfz node_exporter-1.1.2.linux-amd64.tar.gz
cd node_exporter-1.1.2.linux-amd64
nohup ./node_exporter &

查看数据：

$ curl http://localhost:9100/metrics
# HELP go_gc_duration_seconds A summary of the pause duration of garbage collection cycles.
# TYPE go_gc_duration_seconds summary
go_gc_duration_seconds{quantile="0"} 0
go_gc_duration_seconds{quantile="0.25"} 0
go_gc_duration_seconds{quantile="0.5"} 0
...

DCGM Exporter

用于采集 NVIDIA GPU 的数据，以 Docker 镜像[5]运行：

docker run -d --restart=always --gpus all -p 9400:9400 nvidia/dcgm-exporter

查看数据：

$ curl localhost:9400/metrics
# HELP DCGM_FI_DEV_SM_CLOCK SM clock frequency (in MHz).
# TYPE DCGM_FI_DEV_SM_CLOCK gauge
# HELP DCGM_FI_DEV_MEM_CLOCK Memory clock frequency (in MHz).
# TYPE DCGM_FI_DEV_MEM_CLOCK gauge
# HELP DCGM_FI_DEV_MEMORY_TEMP Memory temperature (in C).
...

服务器

Prometheus

配置 ~/prometheus.yml：

global:
  scrape_interval: 15s

scrape_configs:
# Node Exporter
- job_name: node
  static_configs:
  - targets: ['192.167.200.91:9100']
# DCGM Exporter
- job_name: dcgm
  static_configs:
  - targets: ['192.167.200.91:9400']

运行 Docker 镜像：

docker run -d --restart=always \
-p 9090:9090 \
-v ~/prometheus.yml:/etc/prometheus/prometheus.yml \
prom/prometheus

访问 http://localhost:9090/ ：

访问 http://localhost:9090/targets ：

Grafana

运行 Docker 镜像：

docker run -d --restart=always -p 3000:3000 grafana/grafana

访问 http://localhost:3000/ ：

以 admin/admin 登录。

新增数据源

新增 Prometheus：

点击 Save & Test：

导入仪表盘

导入 8919 Node Exporter for Prometheus Dashboard by StarsL.cn[6]：

查看仪表盘：

导入 12239 NVIDIA DCGM Exporter Dashboard by nvidia[7]：

查看仪表盘：

参考

• Start Prometheus[8]
• Prometheus Docs[9]
  • Configuration[10]
  • Node Exporter[11]
  • DCGM Exporter[12]
• Grafana Docs[13]
  • Dashboards[14]
  • Plugins[15]

脚注

[1]Prometheus: https://github.com/prometheus/

[2]Grafana: https://github.com/grafana/

[3]Docker: https://docs.docker.com/

[4]这里下载: https://prometheus.io/download/#node_exporter

[5]Docker 镜像: https://hub.docker.com/r/nvidia/dcgm-exporter

[6]Node Exporter for Prometheus Dashboard by StarsL.cn: https://grafana.com/grafana/dashboards/8919

[7]NVIDIA DCGM Exporter Dashboard by nvidia: https://grafana.com/grafana/dashboards/12239

[8]Start Prometheus: https://github.com/ikuokuo/start-prometheus

[9]Prometheus Docs: https://prometheus.io/docs/

[10]Configuration: https://prometheus.io/docs/prometheus/latest/configuration/configuration/

[11]Node Exporter: https://github.com/prometheus/node_exporter

[12]DCGM Exporter: https://github.com/NVIDIA/gpu-monitoring-tools#dcgm-exporter

[13]Grafana Docs: https://grafana.com/docs/grafana/latest/

[14]Dashboards: https://grafana.com/grafana/dashboards

[15]Plugins: https://grafana.com/grafana/plugins/