[Yandex Cloud documentation](../index.md) > [Yandex MPP Analytics for PostgreSQL](index.md) > Release notes

# Yandex MPP Analytics for PostgreSQL release notes

## May 2026 {#may-2026}

Increased the maximum local disk size per host to 36 TB. Disks of this size are available for Intel Ice Lake (`standard-v3`) configurations with 16 or more cores. To create extra large clusters, you may need to contact [support](https://center.yandex.cloud/support) to reserve the required resources.

## March 2026 {#mar-2026}

* Added separate [roles](security/index.md) for viewing and editing (rescheduling) maintenance without access to the cluster.

## December 2025 {#dec-2025}

* Added [detailed metrics](operations/monitoring.md#hosts) for disk usage on cluster hosts.

## November 2025 {#nov-2025}

Added the `managed-greenplum.restorer` role for restoring clusters from backups. To learn more, see [Access management](security/index.md#managed-greenplum-restorer).

## October 2025 {#oct-2025}

Added new [host classes](concepts/instance-types.md) on AMD Zen 4 and AMD Zen 4 HighFreq platforms.

## July 2025 {#jul-2025}

Added a feature to [transfer cluster logs to Yandex Cloud Logging](operations/mgp-to-cloud-logging.md).


## May 2025 {#may-2025}

Added integration with [Yandex WebSQL](operations/web-sql-query.md).


## April 2025 {#apr-2025}

You can now [manage resource groups](operations/resource-groups.md) using the Yandex Cloud CLI.

## October 2024 {#oct-2024}

* The default [Log statement](concepts/settings-list.md#setting-log-statement) value is now `DDL` rather than `ALL`: Greenplum® only logs SQL commands used to change data structure definitions (such as `CREATE`, `ALTER`, `DROP`, etc.).
* A new RAM limit has been introduced for each Greenplum® process. The limit is calculated using the following formula: `<segment host RAM> ÷ (<Max connections> × <number of segments per host>)`.

## September 2024 {#sep-2024}

* Yandex MPP Analytics for PostgreSQL is now available in the Kazakhstan [region](../overview/concepts/region.md).

* Added the new **Monitoring** and **Network** sections with plots for the respective metrics to the ![monitoring.svg](../_assets/console-icons/display-pulse.svg) **PXF** tab in the management console. For more information, see [Cluster and host state monitoring](operations/monitoring.md).


* Introduced the **Yandex Query access** option in the cluster settings. It enables you to run YQL queries from [Yandex Query](../query/concepts/index.md) to a Yandex MPP Analytics for PostgreSQL database.


## June 2024 {#jun-2024}

You can now [create](operations/index.md#pxf) external S3, JDBC, HDFS, and Hive data sources for accessing external tables.

## February 2024 {#feb-2024}

Added the ability to partially [restore a cluster](operations/cluster-backups.md#restore) from a backup: you can now specify a list of databases or tables to restore. This greatly reduces recovery time and saves cluster resources when restoring small amounts of data.

## January 2024 {#jan-2024}

* A new version is out: Greenplum® 6.25. It features [cold storage integration](tutorials/yezzey.md) tools and a [command center](concepts/command-center.md) by default.
* You can now [configure PXF](operations/pxf/settings.md) for accessing external tables on your own. Previously, users had to contact support to configure PXF.

## Q3 2023 {#q3-2023}

* Added support for [incremental backups](concepts/backup.md).
* Added the option of [committing](../billing/concepts/cvos.md) a specific volume of services. All platform users can get a discount up to 22% starting August 1, 2023. The cost of services will not change while the discount is in effect.
* Added [PXF metrics](metrics.md#managed-greenplum-pxf-metrics) to Yandex Monitoring.
* Added logs for PXF and the Odyssey connection pooler.
* Added [audit logs](at-ref.md) in Yandex Audit Trails.

## Q2 2023 {#q2-2023}

* You can now recover from a backup as of any point in time (Point-in-Time-Recovery, PITR). The feature is enabled by default for new clusters.
* When recovering a cluster, you can now select the number of hosts and segments.
* PXF has been updated to version 6.6.0.
* Logs now support PXF-based events.

## Q4 2022 {#q4-2022}

* You can now [expand a cluster](operations/cluster-expand.md) via the CLI and API by adding new segment hosts with data redistribution across all segments.
* You can now view cluster backups, recover from backups, and modify public access to the cluster via the CLI. For more information, see [YC CLI releases](../cli/release-notes.md#version0.100.0).
* Added the `mdb_toolkit.table_privileges()` function that allows users with the `mdb_admin` role to view the history of granting and revoking privileges in the cluster. For more information, see [this PostgreSQL guide](https://www.postgresql.org/docs/current/infoschema-table-privileges.html).
* Added the `mdb_toolkit.gp_workfile_usage_per_query()` function that allows users with the `mdb_admin` role to view information about active queries from all users. For more information, see [this Greenplum® guide](https://techdocs.broadcom.com/us/en/vmware-tanzu/data-solutions/tanzu-greenplum/7/greenplum-database/ref_guide-gp_toolkit.html#topic34).
* Improved the algorithm for creating clusters via the wizard:
    * The cluster size cannot exceed `<maximum_number_of_hosts> × <maximum_disk_size_per_host>`.
    * You can select groups of dedicated hosts to deploy the cluster.
    * If the amount of data you specified is large, the wizard will prompt you to create a cluster on dedicated hosts.

## Q3 2022 {#q3-2022}

* You can now change the administrator password.
* You can now [configure the DBMS and the connection pooler in Terraform](https://github.com/yandex-cloud/terraform-provider-yandex/blob/master/CHANGELOG.md#0770-july-27-2022).
* Added support for the [postgis](https://techdocs.broadcom.com/us/en/vmware-tanzu/data-solutions/tanzu-greenplum/5/greenplum-database/ref_guide-extensions-postGIS.html) extension.
* Added the ability to edit clusters and cluster settings via the CLI as well as grant access from Yandex Data Transfer. For more information, see [YC CLI releases](../cli/release-notes.md#version0.94.0).
* Updated the master failover mechanism: a former master is returned to a cluster as a replica and is available for another failover right after the first one completes.
* Added the `gp_cancel_backend` function to forcibly stop user queries.
* Added the `gp_terminate_backend` function to forcibly terminate user DB connections.
* Added the ability to expand a cluster with new segment hosts via the API and re-distribute data across all segments.

## Q2 2022 {#q2-2022}

* Added CLI support, with `yc managed-greenplum cluster` and `yc managed-greenplum hosts` commands now available.
* You can now migrate databases from/to Yandex MPP Analytics for PostgreSQL using Yandex Data Transfer. For a list of available targets and sources, see [this guide](../data-transfer/concepts/index.md#connectivity-matrix). This feature is at the [Preview](../overview/concepts/launch-stages.md) stage.
* Added support for the [pgcrypto](https://techdocs.broadcom.com/us/en/vmware-tanzu/data-solutions/tanzu-greenplum/7/greenplum-database/ref_guide-modules-pgcrypto.html) extension.
* [Optimized](https://github.com/wal-g/wal-g/pull/1257) backup creation through special processing of append-only segments.
* Added the [diskquota](https://techdocs.broadcom.com/us/en/vmware-tanzu/data-solutions/tanzu-greenplum/7/greenplum-database/ref_guide-modules-diskquota.html) module that enables you to limit the amount of disk space used by schemas in a database.
* Implemented automatic switchover to a standby master host using the [gpactivatestandby](https://techdocs.broadcom.com/us/en/vmware-tanzu/data-solutions/tanzu-greenplum/7/greenplum-database/utility_guide-ref-gpactivatestandby.html) utility.
* Added the `max_statement_mem` and `log_statement` settings that are available when creating and changing a cluster.
* Added the ability to create a cluster with an odd number of segments.
* Limited the minimum storage size on network SSD disks for a master host to 100 GB.
* Optimized the cleaning process (`VACUUM`):
  * The operation is carried out concurrently across several databases.
  * New tables are processed last.
  * Tables with active locks are excluded.

## Q1 2022 {#q1-2022}

* Yandex MPP Analytics for PostgreSQL is now publicly available. It is now subject to a [Service Level Agreement](https://yandex.com/legal/cloud_sla_greenplum/) (SLA) and [pricing rules](pricing/index.md).
* A new version is out with known bugs fixed: Greenplum® 6.19.
* You can now hide the contents of external tables.
* You can now change the storage size, including for local SSDs.
* Added the `io-optimized` [configurations](concepts/instance-types.md) with an increased GB RAM to vCPU ratio (8:1).

_Greenplum® and Greenplum Database® are registered trademarks or trademarks of Broadcom Inc. in the United States and/or other countries._