Stay organized with collections
Save and categorize content based on your preferences.
When you enable Dataproc cluster caching, the cluster caches
Cloud Storage data frequently accessed by your Spark jobs.
Benefits
Improved performance: Caching can improve job performance by reducing the amount
of time spent retrieving data from storage.
Reduced storage costs: Since hot data is cached on local disk,
fewer API calls are made to storage to retrieve data.
Spark job applicability: When cluster caching is enabled on a cluster,
it applies to all Spark jobs run on the cluster, whether submitted to the
Dataproc service or run independently on the cluster.
Limitations and requirements
Caching applies to Dataproc Spark jobs only.
Only Cloud Storage data is cached.
Caching only applies to clusters that meet the following requirements:
This feature is available in Dataproc on Compute Engine
image versions2.0.72+, 2.1.20+, and 2.2.0+.
Each cluster node must have
local SSDs
attached with the
NVME (Non-Volatile Memory Express)
interface (Persistent Disks (PDs) are not supported). Data is cached on NVME
local SSDs only.
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Hard to understand","hardToUnderstand","thumb-down"],["Incorrect information or sample code","incorrectInformationOrSampleCode","thumb-down"],["Missing the information/samples I need","missingTheInformationSamplesINeed","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2025-09-04 UTC."],[[["\u003cp\u003eEnabling Dataproc cluster caching improves Spark job performance by caching frequently accessed Cloud Storage data on local SSDs, reducing data retrieval time and storage costs.\u003c/p\u003e\n"],["\u003cp\u003eCluster caching applies to all Spark jobs on the cluster, whether submitted to the Dataproc service or run independently, and this applies only to Cloud Storage data.\u003c/p\u003e\n"],["\u003cp\u003eCluster caching is only compatible with clusters meeting specific criteria, such as having one master and \u003ccode\u003en\u003c/code\u003e workers, supported image versions (\u003ccode\u003e2.0.72+\u003c/code\u003e, \u003ccode\u003e2.1.20+\u003c/code\u003e, \u003ccode\u003e2.2.0+\u003c/code\u003e), NVME local SSDs, and the default VM service account.\u003c/p\u003e\n"],["\u003cp\u003eYou can enable cluster caching during Dataproc cluster creation through the Google Cloud console, gcloud CLI, or the Dataproc API, using the property \u003ccode\u003edataproc:dataproc.cluster.caching.enabled=true\u003c/code\u003e.\u003c/p\u003e\n"]]],[],null,["When you enable Dataproc cluster caching, the cluster caches\nCloud Storage data frequently accessed by your Spark jobs.\n\nBenefits\n\n- **Improved performance:** Caching can improve job performance by reducing the amount of time spent retrieving data from storage.\n- **Reduced storage costs:** Since hot data is cached on local disk, fewer API calls are made to storage to retrieve data.\n- **Spark job applicability**: When cluster caching is enabled on a cluster, it applies to all Spark jobs run on the cluster, whether submitted to the Dataproc service or run independently on the cluster.\n\nLimitations and requirements\n\n- Caching applies to Dataproc Spark jobs only.\n- Only Cloud Storage data is cached.\n- Caching only applies to clusters that meet the following requirements:\n - The cluster has one master and `n` workers ([High Availability (HA)](/dataproc/docs/concepts/configuring-clusters/high-availability) and [single node](/dataproc/docs/concepts/configuring-clusters/single-node-clusters) clusters are not supported).\n - This feature is available in Dataproc on Compute Engine [image versions](/dataproc/docs/concepts/versioning/dataproc-version-clusters#supported-dataproc-image-versions) `2.0.72+`, `2.1.20+`, and `2.2.0+`.\n - Each cluster node must have [local SSDs](/dataproc/docs/concepts/compute/dataproc-local-ssds) attached with the [NVME (Non-Volatile Memory Express)](/compute/docs/disks/local-ssd#nvme) interface (Persistent Disks (PDs) are not supported). Data is cached on NVME local SSDs only.\n - The cluster uses the [default VM service account](/dataproc/docs/concepts/configuring-clusters/service-accounts#VM_service_account) for authentication. [Custom VM service accounts](/dataproc/docs/concepts/configuring-clusters/service-accounts#create_a_cluster_with_a_custom_vm_service_account) are not supported.\n\nEnable cluster caching\n\nYou can enable cluster caching when you create a Dataproc cluster\nusing the Google Cloud console, Google Cloud CLI, or the Dataproc API. \n\nGoogle Cloud console\n\n- Open the Dataproc [**Create a cluster on Compute Engine**](https://console.cloud.google.com/dataproc/clustersAdd) page in the Google Cloud console.\n- The **Set up cluster** panel is selected. In the **Spark performance enhancements** section, select **Enable Google Cloud Storage caching**.\n- After confirming and specifying cluster details in the cluster create panels, click **Create**.\n\ngcloud CLI\n\nRun the [gcloud dataproc clusters create](/sdk/gcloud/reference/dataproc/clusters/create)\ncommand locally in a terminal window or in\n[Cloud Shell](https://console.cloud.google.com/?cloudshell=true)\nusing the `dataproc:dataproc.cluster.caching.enabled=true`\n[cluster property](/dataproc/docs/concepts/configuring-clusters/cluster-properties#dataproc_service_properties_table).\n\nExample: \n\n```\ngcloud dataproc clusters create CLUSTER_NAME \\\n --region=REGION \\\n --properties dataproc:dataproc.cluster.caching.enabled=true \\\n --num-master-local-ssds=2 \\\n --master-local-ssd-interface=NVME \\\n --num-worker-local-ssds=2 \\\n --worker-local-ssd-interface=NVME \\\n other args ...\n \n```\n\nREST API\n\nSet [SoftwareConfig.properties](/static/dataproc/docs/reference/rest/v1/ClusterConfig#SoftwareConfig.FIELDS.properties)\nto include the `\"dataproc:dataproc.cluster.caching.enabled\": \"true\"`\n[cluster property](/dataproc/docs/concepts/configuring-clusters/cluster-properties#dataproc_service_properties_table)\nas part of a\n[clusters.create](/dataproc/docs/reference/rest/v1/projects.regions.clusters/create)\nrequest."]]
