共计 1700 个字符,预计需要花费 5 分钟才能阅读完成。
Prometheus + Grafana 疾速上手,监控主机的 CPU, GPU, MEM, IO 等状态。
前提
- Docker
客户端
Node Exporter
用于采集 UNIX 内核主机的数据,这里下载并解压:
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
拜访 http://localhost:9090/:
拜访 http://localhost:9090/targets:
Grafana
运行 Docker 镜像:
docker run -d --restart=always -p 3000:3000 grafana/grafana
拜访 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:
查看仪表盘:
导入 12239
NVIDIA DCGM Exporter Dashboard by nvidia:
查看仪表盘:
参考
- Start Prometheus
-
Prometheus Docs
- Configuration
- Node Exporter
- DCGM Exporter
-
Grafana Docs
- Dashboards
- Plugins
GoCoding 集体实际的教训分享,可关注公众号!
正文完