Yandex Data Processing release notes

Q4 2025Q4 2025

• Added a new host class: AMD Zen 4.
• Added a dedicated IP address for UI Proxy: 158.160.167.170/32.
• You can now specify a separate service account to manage VMs of an autoscaling subcluster when creating a cluster, creating an autoscaling cluster, or updating a cluster.

Q3 2025Q3 2025

In image 2.2.9 (beta), Apache Spark™ is updated to version 3.5.6.

Q2 2025Q2 2025

• Added the OS Login option to use when creating a cluster. This option enables OS Login access to all hosts you create in the cluster.

• Added support for environment variables:

  • HADOOP_HEAPSIZE_MIN and HADOOP_HEAPSIZE_MAX for the hadoop service:
    • hadoop.env:HADOOP_HEAPSIZE_MIN
    • hadoop.env:HADOOP_HEAPSIZE_MAX
  • HADOOP_HEAPSIZE for hive (available only for 2.0 images): hive.env:HADOOP_HEAPSIZE.

Q1 2025Q1 2025

In 2.2.X images, Java version updated to 11.

Q4 2024Q4 2024

• Added environment selection (PRODUCTION / PRESTABLE) during cluster creation and modification.
• In 2.2.X images, Python version updated to 3.1.

Q3 2024Q3 2024

• Apache Hive™ Metastore cluster functionality has been integrated in Yandex MetaData Hub. For more information about Apache Hive™ Metastore clusters, see the Yandex MetaData Hub documentation.
• In 2.1.X and 2.2.X images, Conda now uses Mamba as its default solver.

Q2 2024Q2 2024

A stable image version line 2.1 is now available. This update enables cluster creation with newer runtime versions: Spark 3.3.2 and Hadoop 3.3.2.

Q2 2023Q2 2023

Added support for creating Apache Hive™ Metastore clusters. This feature is currently in Preview.

Q3 2022Q3 2022

• Added support for new configuration settings in the DataprocCreateClusterOperator Airflow operator.
• Added cpu-optimized host classes configured with 2GB RAM per 1 vCPU core. The new configurations are exclusively available for Intel Ice Lake processors.
• Published a guide for using initialization scripts to configure GeeseFS.

Q2 2022Q2 2022

• Image version 2.1 is now available.
• Added support for public internet access across all subcluster types.
• Lightweight Spark support is now available starting with image version 2.0.39. You can now create a cluster without data storage subclusters because YARN and SPARK services are no longer dependent on HDFS.
• Added support for initialization scripts in the CLI.

Q1 2022Q1 2022

• You can now create clusters using non-replicated network drives up to 8 TB in size. Non-replicated drives have a simpler architecture than network SSD storage, resulting in significantly higher performance.
• Added support for job cancellation.
• Added the build number in Yandex Data Processing image version.
• Spark and PySpark jobs now accept packages, repositories, and exclude_packages parameters. You can use these parameters to download additional dependencies and packages from third-party repositories.