このページでは、Google Kubernetes Engine（GKE）で実行されている Ray クラスタのログと指標を収集するように Google Kubernetes Engine（GKE）を構成する方法と、Cloud Logging と Cloud Monitoring で Ray のログと指標を表示する方法について説明します。
Ray クラスタのロギングを有効にする前に、既存の GKE クラスタでシステムとワークロードのロギングを有効にする必要があります。
既存の GKE クラスタで Ray クラスタのログ収集を有効にすると、GKE は既存の Ray Pod ではなく、新しく作成された Ray Pod からのみログを収集します。
GKE Standard クラスタの場合、Ray クラスタの指標の収集を有効にするには、Google Cloud Managed Service for Prometheus を有効にする必要があります。Autopilot クラスタの場合、Google Cloud Managed Service for Prometheus はデフォルトで有効になっています。
Ray クラスタ内の Ray コンテナで ray-logs という名前のボリュームを指定しないでください。指定すると、GKE はログを収集しません。
Ray クラスタのログ収集を有効にする
Ray クラスタのログ収集は、新規または既存の GKE Autopilot または Standard クラスタで有効にできます。GKE が Ray クラスタから収集する Ray ログは、コンテナログとして分類されます。これには、Ray クラスタ ヘッダーノードとワーカーノードで生成されたすべてのログが含まれます。
Ray クラスタに対するログ収集を有効にするには、 Google Cloud コンソールまたは gcloud CLI を使用します。
コンソール
Google Cloud コンソールで [Google Kubernetes Engine] ページに移動します。
Google Cloud Managed Service for Prometheus には、事前構成された Ray on GKE の概要のダッシュボードが用意されています。このダッシュボードには、主要な Ray 指標が一元的に表示されます。これは、GKE で Ray クラスタのモニタリングを迅速に開始するうえでおすすめの方法です。
[[["わかりやすい","easyToUnderstand","thumb-up"],["問題の解決に役立った","solvedMyProblem","thumb-up"],["その他","otherUp","thumb-up"]],[["わかりにくい","hardToUnderstand","thumb-down"],["情報またはサンプルコードが不正確","incorrectInformationOrSampleCode","thumb-down"],["必要な情報 / サンプルがない","missingTheInformationSamplesINeed","thumb-down"],["翻訳に関する問題","translationIssue","thumb-down"],["その他","otherDown","thumb-down"]],["最終更新日 2025-07-30 UTC。"],[],[],null,["# Collect and view logs and metrics for Ray clusters on Google Kubernetes Engine (GKE)\n\n[Autopilot](/kubernetes-engine/docs/concepts/autopilot-overview) [Standard](/kubernetes-engine/docs/concepts/choose-cluster-mode)\n\n*** ** * ** ***\n\nThis page shows how to configure Google Kubernetes Engine (GKE) to collect logs\nand metrics for Ray clusters running on Google Kubernetes Engine (GKE), plus how to\nview Ray logs and metrics in Cloud Logging and Cloud Monitoring.\n\nFor more\ninformation on Ray and KubeRay, see\n[Ray on Google Kubernetes Engine (GKE) overview](/kubernetes-engine/docs/add-on/ray-on-gke/concepts/overview).\n\nBefore you begin\n----------------\n\nBefore you start, make sure that you have performed the following tasks:\n\n- Enable the Google Kubernetes Engine API.\n[Enable Google Kubernetes Engine API](https://console.cloud.google.com/flows/enableapi?apiid=container.googleapis.com)\n- If you want to use the Google Cloud CLI for this task, [install](/sdk/docs/install) and then [initialize](/sdk/docs/initializing) the gcloud CLI. If you previously installed the gcloud CLI, get the latest version by running `gcloud components update`. **Note:** For existing gcloud CLI installations, make sure to set the `compute/region` [property](/sdk/docs/properties#setting_properties). If you use primarily zonal clusters, set the `compute/zone` instead. By setting a default location, you can avoid errors in the gcloud CLI like the following: `One of [--zone, --region] must be supplied: Please specify location`. You might need to specify the location in certain commands if the location of your cluster differs from the default that you set.\n\n\u003c!-- --\u003e\n\n- [Enable the Ray operator for Google Kubernetes Engine (GKE)](/kubernetes-engine/docs/add-on/ray-on-gke/how-to/enable-ray-on-gke).\n\n### Requirements and limitations\n\n- You must enable system and workload logging on an existing GKE cluster before you enable log collection for Ray clusters.\n- If you enable log collection for Ray clusters on an existing GKE cluster, GKE only collects logs from newly created Ray Pods, not from existing Ray Pods.\n- For Standard GKE clusters, you must enable Google Cloud Managed Service for Prometheus to enable metrics collection for Ray clusters. For Autopilot clusters, Google Cloud Managed Service for Prometheus is enabled by default.\n- You must **not** specify a volume named `ray-logs` in any Ray container in the Ray cluster. Otherwise, GKE won't collect logs.\n\nEnable log collection for a Ray cluster\n---------------------------------------\n\nYou can enable log collection for Ray clusters with new or existing\nAutopilot or Standard GKE clusters. The Ray\nlogs that GKE collects from Ray clusters are classified as\ncontainer logs. This includes all logs produced by the Ray cluster header and\nworker nodes.\n\nYou can enable log collection for Ray clusters using the Google Cloud console\nor the gcloud CLI. \n\n### Console\n\n1. Go to the **Google Kubernetes Engine** page in the Google Cloud console.\n\n [Go to Google Kubernetes Engine](https://console.cloud.google.com/kubernetes/list)\n2. Click add_box **Create** then in the Standard or Autopilot section, click **Configure**.\n\n3. From the navigation pane, under **Cluster** , click **Features**.\n\n4. In the **Operations** section, ensure the **System and Workloads**\n checkbox is selected.\n\n5. In the **AI and Machine Learning** section, select\n **Enable Ray Operator** and then select **Enable log collection for\n Ray clusters**.\n\n6. Click **Create**.\n\nFor Standard clusters, you must also enable\nGoogle Cloud Managed Service for Prometheus.\n\n### gcloud\n\nCreate a cluster using the `--addons=RayOperator` option and the\n`--enable-ray-cluster-logging` option: \n\n gcloud container clusters create \u003cvar translate=\"no\"\u003eCLUSTER_NAME\u003c/var\u003e \\\n --location=\u003cvar translate=\"no\"\u003eLOCATION\u003c/var\u003e \\\n --addons=RayOperator \\\n --enable-ray-cluster-logging\n\nReplace the following:\n\n- \u003cvar translate=\"no\"\u003eCLUSTER_NAME\u003c/var\u003e: the name of the new cluster.\n- \u003cvar translate=\"no\"\u003eLOCATION\u003c/var\u003e: the location of the new cluster, for example, us-central1.\n\nYou can enable log collection for Ray clusters on an existing cluster by\nusing the\n[`gcloud container clusters update`](/sdk/gcloud/reference/container/clusters/update)\ncommand with the `--addons=RayOperator` option and the\n`--enable-ray-cluster-logging` option.\n| **Note:** You might observe that, in GKE, the Ray Operator collects logs from Ray head and worker Pods (standard output and standard error) even when the **Enable log collection for Ray clusters** option is not selected. This behavior is expected because GKE, by default, automatically collects all workload logs written to standard output or standard error. The **Enable log collection for Ray clusters** checkbox specifically controls the collection of additional Ray-specific logs, separate from these default workload logs. To manage which logs are sent to Cloud Logging by default and to reduce logging volume, refer to the [About GKE logs](/kubernetes-engine/docs/concepts/about-logs#what_logs) page.\n\nView Ray logs\n-------------\n\nYou can view logs collected from Ray clusters running on GKE\nusing Logging.\n\n1. Go to the **Cloud Logging** page in the Google Cloud console.\n\n [Go to Cloud Logging](https://console.cloud.google.com/logs)\n2. Open the query editor and paste your expression into the query editor\n\n3. Click **Run query**\n\nYou can use the following examples queries in the Logs Explorer:\n\nEnable metrics collection for a Ray cluster\n-------------------------------------------\n\nYou can enable metrics collection for Ray clusters with new or existing\nAutopilot or Standard GKE clusters.\n\nAfter you enable metrics collection for Ray clusters, GKE\ncollects metrics from existing Ray clusters and new Ray clusters.\nGKE collects all system metrics exported by Ray in Prometheus\nformat.\n\nYou can enable metrics collection for Ray clusters using the\nGoogle Cloud console or the gcloud CLI. \n\n### Console\n\n1. Go to the **Google Kubernetes Engine** page in the Google Cloud console.\n\n [Go to Google Kubernetes Engine](https://console.cloud.google.com/kubernetes/list)\n2. Click add_box **Create** then in the Standard or Autopilot section, click **Configure**.\n\n3. From the navigation pane, under **Cluster** , click **Features**.\n\n4. In the **Operations** section, ensure the **System and Workloads**\n checkbox is selected.\n\n5. In the **AI and Machine Learning** section, select\n **Enable Ray Operator** and then select **Enable metrics collection for\n Ray clusters**.\n\n6. Click **Create**.\n\nFor Standard clusters, you must also enable\nGoogle Cloud Managed Service for Prometheus.\n\n### gcloud\n\nCreate a cluster using the `--addons=RayOperator` option and the\n`--enable-ray-cluster-monitoring` option: \n\n gcloud container clusters create \u003cvar translate=\"no\"\u003eCLUSTER_NAME\u003c/var\u003e \\\n --location=\u003cvar translate=\"no\"\u003eLOCATION\u003c/var\u003e \\\n --addons=RayOperator \\\n --enable-ray-cluster-monitoring\n\nReplace the following:\n\n- \u003cvar translate=\"no\"\u003eCLUSTER_NAME\u003c/var\u003e: the name of the new cluster.\n- \u003cvar translate=\"no\"\u003eLOCATION\u003c/var\u003e: the location of the new cluster, for example, us-central1.\n\nYou can enable log collection for Ray clusters on an existing cluster by\nusing the\n[`gcloud container clusters update`](/sdk/gcloud/reference/container/clusters/update)\ncommand with the `--addons=RayOperator` option and the\n`--enable-ray-cluster-monitoring` option.\n\nView Ray metrics\n----------------\n\nGoogle Cloud Managed Service for Prometheus provides a pre-configured\n**Ray on GKE Overview** dashboard that offers a centralized view\nof key Ray metrics. This is the recommended way\nto quickly get started with monitoring your Ray clusters on GKE.\n\n[Go to Ray on GKE Overview dashboard](https://console.cloud.google.com/monitoring/dashboards/integration/kuberay.ray-overview)\n\nThe dashboard is automatically populated when you [enable\nmetrics collection](/kubernetes-engine/docs/add-on/ray-on-gke/how-to/collect-view-logs-metrics#enable-metrics-collection) for your Ray cluster.\n\nAlternatively, if you want to explore individual metrics collected from Ray\nclusters running on GKE, follow these steps:\n\n1. Go to the **Metrics Explorer** page in the Google Cloud console.\n\n [Go to Metrics Explorer](https://console.cloud.google.com/monitoring/metrics-explorer)\n2. In the **Select a metric** field, you can search for Ray-specific metrics.\n These metrics are typically prefixed with `prometheus/ray_`. Examples include\n `prometheus/ray_worker_cpu_seconds_total` or `prometheus/ray_memory_bytes_max`.\n\n3. You can further refine your search by selecting the appropriate resource type\n (for example, `k8s_pod`, `k8s_container`) and filtering by labels relevant to\n your Ray cluster (for example, `ray.io/cluster`).\n\nWhat's next\n-----------\n\n- Learn about [Ray on Kubernetes](https://docs.ray.io/en/latest/cluster/kubernetes/index.html).\n- Explore the [KubeRay documentation](https://docs.ray.io/en/latest/cluster/kubernetes/getting-started.html)."]]
