部署prometheus监控vllm

 

1、访问官网下载：

wget https://github.com/prometheus/prometheus/releases/download/v3.5.0/prometheus-3.5.0.linux-amd64.tar.gz

2、安装：

mkdir /usr/local/prometheus
tar -zxf prometheus-3.5.0.linux-amd64.tar.gz
cp -r prometheus-3.5.0.linux-amd64/* /usr/local/prometheus/

3、配置服务项：
vim /etc/systemd/system/prometheus.service ，内容如下：

[Unit]
Description="prometheus"
Documentation=https://prometheus.io/
After=network.target
[Service]
Type=simple
ExecStart=/usr/local/prometheus/prometheus  --config.file=/usr/local/prometheus/prometheus.yml --storage.tsdb.path=/usr/local/prometheus/data --web.enable-lifecycle --enable-feature=remote-write-receiver --query.lookback-delta=2m --web.enable-admin-api
ExecReload=/bin/kill -HUP $MAINPID
Restart=on-failure
SuccessExitStatus=0
LimitNOFILE=65536
StandardOutput=syslog
StandardError=syslog
SyslogIdentifier=prometheus
[Install]
WantedBy=multi-user.target

启动服务：

systemctl start prometheus
systemctl enable prometheus
systemctl status prometheus

4、访问http://127.0.0.1:9000可以打开页面

 

编辑配置文件，增加对vllm的监控，如下：

vim /usr/local/prometheus/prometheus.yml

scrape_configs:
  # The job name is added as a label `job=<job_name>` to any timeseries scraped from this config.
  - job_name: "vllm_job"

    # metrics_path defaults to '/metrics'
    # scheme defaults to 'http'.

    static_configs:
      - targets: ["10.0.11.5:801"]                                                                                                                                                                                                                                        
       # The label name is added as a label `label_name=<label_value>` to any timeseries scraped from this config.
        labels:
          app: "vllm"
~                                

保存后system reload prometheus

在Status-Target health页面下可以看到监控目标服务器：

 在Query页面，可以查询vllm的各个监控指标