Use hierarchical namespace enabled buckets for Hadoop workloads
Stay organized with collections
Save and categorize content based on your preferences.
This page describes how to use hierarchical namespace enabled buckets for Hadoop workloads.
Overview
When using a Cloud Storage bucket with hierarchical namespace, you can configure the Cloud Storage connector to use the rename folder operation for workloads like Hadoop, Spark, Hive.
In a bucket without hierarchical namespace, a rename operation in Hadoop, Spark,
and Hive involves multiple object copy and delete jobs, impacting
performance and consistency. Renaming a folder using the Cloud Storage
connector optimizes performance and ensures consistency, when handling folders
with a large number of objects.
Before you begin
To use features of hierarchical namespace buckets, use the following Cloud Storage
connector versions:
fs.gs.hierarchical.namespace.folders.enable is used to enable the hierarchical namespace on a bucket
fs.gs.http.read-timeout is the maximum time allowed, in milliseconds, to read data from an established connection. This is an optional setting.
Compatibility with Cloud Storage connector version 3.0.0 or versions older than 2.2.23
Using the Cloud Storage connector version 3.0.0 or versions older than 2.2.23 or disabling folder operations for hierarchical namespace can lead to the following limitations:
Inefficient folder renames: Folder rename operations in Hadoop happen
using object-level copy and delete operations, which is slower and less
efficient than the dedicated rename folder operation.
Accumulation of empty folders: Folders are not deleted
automatically, leading to the accumulation of empty folders in your bucket.
Accumulation of empty folders can have the following impact:
Increase storage costs if not deleted explicitly.
Slow down the list operations and increase the risk of list operation
timeouts.
Compatibility issues: Mixing the usage of older and newer connector
versions, or enabling and disabling folder operations, can lead to
compatibility issues, when renaming folders. Consider the following scenario
which uses a combination of connector versions:
Use the Cloud Storage connector version older than 2.2.23 to
perform the following tasks:
Write objects under the folder foo/.
Rename the folder foo/ to bar/. The rename operation copies and
deletes the objects under foo/ but does not delete the empty foo/
folder.
Use the Cloud Storage connector version 2.2.23 with the
folder operations settings enabled to rename the folder bar/ to foo/.
The connector version 2.2.23, with the folder operation enabled,
detects the existing foo/ folder, causing the rename operation to
fail. The older connector version, did not delete the foo/ folder as
the folder operation was disabled.
If you're new to Google Cloud, create an account to evaluate how
Cloud Storage performs in real-world
scenarios. New customers also get $300 in free credits to run, test, and
deploy workloads.
[[["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 2025-08-07 UTC."],[],[],null,["# Use hierarchical namespace enabled buckets for Hadoop workloads\n\nThis page describes how to use [hierarchical namespace](/storage/docs/hns-overview) enabled buckets for Hadoop workloads.\n\nOverview\n--------\n\nWhen using a Cloud Storage bucket with hierarchical namespace, you can configure the [Cloud Storage connector](/dataproc/docs/concepts/connectors/cloud-storage) to use the [rename folder](/storage/docs/rename-hns-folders) operation for workloads like Hadoop, Spark, Hive.\n\nIn a bucket without hierarchical namespace, a rename operation in Hadoop, Spark,\nand Hive involves multiple object copy and delete jobs, impacting\nperformance and consistency. Renaming a folder using the Cloud Storage\nconnector optimizes performance and ensures consistency, when handling folders\nwith a large number of objects.\n\nBefore you begin\n----------------\n\nTo use features of hierarchical namespace buckets, use the following Cloud Storage\nconnector versions:\n\n- 2.2.23 or later (if you are using version 2.x.x)\n- 3.0.1 or later (if you are using version 3.x.x)\n\nOlder connector versions (3.0.0 and older than 2.2.23) have limitations. For more information about the limitations, see [Compatibility with\nCloud Storage connector version 3.0.0 or versions older than\n2.2.23](/storage/docs/hns-hadoop-workloads#backwards-compatibility-connector).\n\nEnable the Cloud Storage connector on a cluster\n-----------------------------------------------\n\nThis section describes how to enable the Cloud Storage connector on a Dataproc cluster and a self-managed Hadoop cluster. \n\n### Dataproc\n\nYou can use the Google Cloud CLI to create a Dataproc cluster and enable the Cloud Storage connector to perform the folder operations.\n\n1. Create a Dataproc cluster using the following command:\n\n \u003cbr /\u003e\n\n ```\n gcloud dataproc clusters create CLUSTER_NAME\n --properties=core:fs.gs.hierarchical.namespace.folders.enable=true,\n core:fs.gs.http.read-timeout=30000\n \n ```\n\n \u003cbr /\u003e\n\n Where:\n - \u003cvar translate=\"no\"\u003eCLUSTER_NAME\u003c/var\u003e is the name of the cluster. For example, `my-cluster`\n - `fs.gs.hierarchical.namespace.folders.enable` is used to enable the hierarchical namespace on a bucket.\n - `fs.gs.http.read-timeout` is the maximum time allowed, in milliseconds, to read data from an established connection. This is an optional setting.\n\n | **Note:** If you are using the Cloud Storage connector version 3.0.0 or a version older than 2.2.23, the configuration setting `fs.gs.hierarchical.namespace.folders.enable` is not supported and results in an error if included.\n\n### Self-managed Hadoop\n\nYou can enable the Cloud Storage connector on your self-managed Hadoop cluster to perform the folder operations.\n\n1. Add the following to core-site.xml configuration file:\n\n \u003cbr /\u003e\n\n ```\n \u003cproperty\u003e\n \u003cname\u003efs.gs.hierarchical.namespace.folders.enable\u003c/name\u003e\n \u003cvalue\u003etrue\u003c/value\u003e\n \u003c/property\u003e\n \u003cproperty\u003e\n \u003cname\u003efs.gs.http.read-timeout\u003c/name\u003e\n \u003cvalue\u003e30000\u003c/value\u003e\n \u003c/property\u003e\n \n ```\n\n \u003cbr /\u003e\n\n Where:\n - `fs.gs.hierarchical.namespace.folders.enable` is used to enable the hierarchical namespace on a bucket\n - `fs.gs.http.read-timeout` is the maximum time allowed, in milliseconds, to read data from an established connection. This is an optional setting.\n\n | **Note:** If you are using the Cloud Storage connector version 3.0.0 or a version older than 2.2.23, the configuration setting `fs.gs.hierarchical.namespace.folders.enable` is not supported and results in an error if included.\n\nCompatibility with Cloud Storage connector version 3.0.0 or versions older than 2.2.23\n--------------------------------------------------------------------------------------\n\nUsing the Cloud Storage connector version 3.0.0 or versions older than 2.2.23 or disabling folder operations for hierarchical namespace can lead to the following limitations:\n\n- **Inefficient folder renames** : Folder rename operations in Hadoop happen\n using object-level copy and delete operations, which is slower and less\n efficient than the dedicated `rename folder` operation.\n\n- **Accumulation of empty folders**: Folders are not deleted\n automatically, leading to the accumulation of empty folders in your bucket.\n Accumulation of empty folders can have the following impact:\n\n - Increase storage costs if not deleted explicitly.\n - Slow down the list operations and increase the risk of list operation\n timeouts.\n\n | **Note:** To reduce the risk of list operation timeouts, configure the `fs.gs.http.read-timeout` timeout value to `30000` milliseconds. To configure timeout settings, refer to the instructions for [Dataproc](/storage/docs/hns-hadoop-workloads#dataproc) or [Self-managed Hadoop](/storage/docs/hns-hadoop-workloads#self-managed-hadoop), depending on which one you are using.\n- **Compatibility issues**: Mixing the usage of older and newer connector\n versions, or enabling and disabling folder operations, can lead to\n compatibility issues, when renaming folders. Consider the following scenario\n which uses a combination of connector versions:\n\n 1. Use the Cloud Storage connector version older than 2.2.23 to\n perform the following tasks:\n\n 1. Write objects under the folder `foo/`.\n 2. Rename the folder `foo/` to `bar/`. The rename operation copies and deletes the objects under `foo/` but does not delete the empty `foo/` folder.\n 2. Use the Cloud Storage connector version 2.2.23 with the\n folder operations settings enabled to rename the folder `bar/` to `foo/`.\n\n The connector version 2.2.23, with the folder operation enabled,\n detects the existing `foo/` folder, causing the rename operation to\n fail. The older connector version, did not delete the `foo/` folder as\n the folder operation was disabled.\n\nWhat's next\n-----------\n\n- [Create buckets with hierarchical namespace enabled](/storage/docs/create-hns-bucket).\n- [Create and manage folders](/storage/docs/create-folders).\n\nTry it for yourself\n-------------------\n\n\nIf you're new to Google Cloud, create an account to evaluate how\nCloud Storage performs in real-world\nscenarios. New customers also get $300 in free credits to run, test, and\ndeploy workloads.\n[Try Cloud Storage free](https://console.cloud.google.com/freetrial)"]]
