Monitoring 에이전트는 CPU 사용률이나 CPU 사용량을 보고할 수 있지만 두 측정항목 모두 보고할 수는 없습니다.
이는 collectd의 알려진 제한사항입니다.
ReportByCpu가 true이면 ValuesPercentage 필드는 사용률이나 사용량이 보고되는지 확인합니다. ReportByCpu가 false이면 에이전트가 CPU 사용률을 보고합니다.
[[["이해하기 쉬움","easyToUnderstand","thumb-up"],["문제가 해결됨","solvedMyProblem","thumb-up"],["기타","otherUp","thumb-up"]],[["이해하기 어려움","hardToUnderstand","thumb-down"],["잘못된 정보 또는 샘플 코드","incorrectInformationOrSampleCode","thumb-down"],["필요한 정보/샘플이 없음","missingTheInformationSamplesINeed","thumb-down"],["번역 문제","translationIssue","thumb-down"],["기타","otherDown","thumb-down"]],["최종 업데이트: 2025-09-04(UTC)"],[],[],null,["# Configure the Monitoring agent\n\nThis page describes how to configure which built-in metrics the\nMonitoring agent collects from a Linux VM.\nThe content on this page is intended for system administrators who want to\nmodify the default configuration of the Monitoring agent.\n\nYou cannot configure the Monitoring agent when it is installed on a Windows VM.\n\nThis page doesn't describe how to configure the Monitoring agent to monitor\nthird-party applications or how to use the agent to collect user-defined\nmetrics. For information on those topics, see the following pages:\n\n- [Monitoring third party applications](/monitoring/agent/plugins)\n- [User-defined metrics from the agent](/monitoring/agent/custom-metrics-agent)\n\nBefore you begin\n----------------\n\nIf you haven't installed the Monitoring agent on a VM,\nsee [Install the Monitoring agent](/stackdriver/docs/solutions/agents/monitoring/installation#joint-install).\n\nChanging the default Monitoring agent configuration\n---------------------------------------------------\n\nTo change the configuration of the Monitoring agent, do the following:\n\n1. Edit the Monitoring agent configuration file `/etc/stackdriver/collectd.conf`.\n\n2. Restart the Monitoring agent by running the following command:\n\n sudo service stackdriver-agent restart\n\nStopping the collection of process metrics\n------------------------------------------\n\nFor information about how to stop the collection of\nprocess metrics---the metrics with the prefix\n`agent.googleapis.com/processes`---see\n[Disabling process metrics on Linux VMs](/monitoring/agent/process-metrics#disable-on-linux).\n\nConfiguring reported CPU metrics\n--------------------------------\n\nBy default, the Monitoring agent reports the total CPU utilization by state\nfor all CPUs on the VM by using the\n[`agent.googleapis.com/cpu/utilization`](/monitoring/api/metrics_agent#cpu/utilization) metric type.\nThis behavior is specified in the configuration file by the following\nsettings:\n\n- `ValuesPercentage true`\n- `ReportByCpu false`\n\nYou must modify these settings if you want the agent to report the\nutilization per CPU, or if you want it to report CPU usage time,\n[`agent.googleapis.com/cpu/usage_time`](/monitoring/api/metrics_agent#cpu/usage_time), metrics:\n\n- To configure the Monitoring agent to report the\n [`agent.googleapis.com/cpu/utilization`](/monitoring/api/metrics_agent#cpu/utilization) per CPU,\n do the following:\n\n - Leave `ValuesPercentage` at the default value of `true`.\n - Set `ReportByCpu` to `true`.\n- To configure the Monitoring agent to report the\n [`agent.googleapis.com/cpu/usage_time`](/monitoring/api/metrics_agent#cpu/usage_time) per CPU, do the following:\n\n - Set `ValuesPercentage` to `false`.\n - Set `ReportByCpu` to `true`.\n\nThe Monitoring agent can either report the CPU utilization or the CPU usage,\nbut not both metrics.\nThis is a known limitation of [collectd](https://collectd.org/wiki/index.php/Main_Page).\nWhen `ReportByCpu` is `true`, the `ValuesPercentage` field determines\nif the utilization or usage is reported. When `ReportByCpu` is `false`,\nthe agent reports the CPU utilization.\n\nTroubleshooting\n---------------\n\nIf you're having trouble installing or running the agent, see\n[Troubleshooting agent installation](/stackdriver/docs/solutions/agents/monitoring/troubleshooting)."]]
