云监控服务 CES-安装GPU指标与RAID指标采集插件(Linux):操作步骤(新版本)

时间:2023-11-01 16:19:48

操作步骤(新版本)

  1. 使用root帐号,登录E CS
    • 若要监控BMS的软RAID指标,请登录BMS。
    • 以下以安装GPU插件为例,安装监控软RAID插件类似。
  2. 执行以下命令,进入Agent安装路径。

    cd /usr/local/uniagent/extension/install/telescope

  3. 执行以下命令,创建plugins文件夹。

    mkdir plugins

  4. 执行以下命令,进入plugins文件夹。

    cd plugins

  5. 执行如下命令,下载采集插件脚本(以下以GPU插件为例)。

    wget https://telescope.obs.cn-north-1.myhuaweicloud.com/gpu_collector

    表1 获取插件采集安装包

    名称

    下载路径

    Linux 64位GPU指标采集插件安装包

    华北-北京一https://obs.cn-north-1.myhuaweicloud.com/uniagent-cn-north-1/extension/gpu/gpu_collector

    华北-北京四https://obs.cn-north-4.myhuaweicloud.com/uniagent-cn-north-4/extension/gpu/gpu_collector

    华南-广州:https://obs.cn-south-1.myhuaweicloud.com/uniagent-cn-south-1/extension/gpu/gpu_collector

    华东-上海一:https://obs.cn-east-3.myhuaweicloud.com/uniagent-cn-east-3/extension/gpu/gpu_collector

    华东-上海二:https://obs.cn-east-2.myhuaweicloud.com/uniagent-cn-east-2/extension/gpu/gpu_collector

    中国-香港:https://obs.ap-southeast-1.myhuaweicloud.com/uniagent-ap-southeast-1/extension/gpu/gpu_collector

    亚太-曼谷:https://obs.ap-southeast-2.myhuaweicloud.com/uniagent-ap-southeast-2/extension/gpu/gpu_collector

    亚太-新加坡:https://obs.ap-southeast-3.myhuaweicloud.com/uniagent-ap-southeast-3/extension/gpu/gpu_collector

    Linux 64位RAID指标采集插件安装包

    华北-北京一https://obs.cn-north-1.myhuaweicloud.com/uniagent-cn-north-1/extension/raid/raid_monitor.sh

    华北-北京四https://obs.cn-north-4.myhuaweicloud.com/uniagent-cn-north-4/extension/raid/raid_monitor.sh

    华南-广州:https://obs.cn-south-1.myhuaweicloud.com/uniagent-cn-south-1/extension/raid/raid_monitor.sh

    华东-上海二:https://obs.cn-east-2.myhuaweicloud.com/uniagent-cn-east-2/extension/raid/raid_monitor.sh

    中国-香港:https://obs.ap-southeast-1.myhuaweicloud.com/uniagent-ap-southeast-1/extension/raid/raid_monitor.sh

    亚太-曼谷:https://obs.ap-southeast-2.myhuaweicloud.com/uniagent-ap-southeast-2/extension/raid/raid_monitor.sh

  6. 执行如下命令,添加脚本执行权限。

    chmod 755 gpu_collector

  7. 执行如下命令,新建conf.json文件并添加配置内容,配置插件路径和指标采集周期crontime(单位:秒)。

    vi conf.json

    GPU指标插件配置

    {    "plugins": [      {        "path": "/usr/local/uniagent/extension/install/telescope/plugins/gpu_collector",        "crontime": 60      }  ] }

    RAID指标插件配置

    {    "plugins": [      {        "path": "/usr/local/uniagent/extension/install/telescope/plugins/raid_monitor.sh",        "crontime": 60      }  ] }
    • path路径后的参数为gpu_collector和raid_monitor.sh分别为GPU插件和RAID插件配置内容。
    • 插件采集周期为60s,若采集周期配置错误,会导致指标采集异常。
    • 插件路径path请勿私自修改,否则指标采集异常。
  8. 打开“/usr/local/uniagent/extension/install/telescope/bin”路径下的conf_ces.json文件,新增配置项"EnablePlugin":true,开启插件采集开关。
    {        "Endpoint": "所在区域地址,默认无需修改",    "EnablePlugin": true}
  9. 执行如下命令,重启Agent。

    ps -ef | grep telescope | grep -v grep | awk '{print $2}' | xargs kill -9

support.huaweicloud.com/usermanual-ces/ces_01_0038.html