Yandex Cloud
Search
Contact UsTry it for free
  • Customer Stories
  • Documentation
  • Blog
  • All Services
  • System Status
  • Marketplace
    • Featured
    • Infrastructure & Network
    • Data Platform
    • AI for business
    • Security
    • DevOps tools
    • Serverless
    • Monitoring & Resources
  • All Solutions
    • By industry
    • By use case
    • Economics and Pricing
    • Security
    • Technical Support
    • Start testing with double trial credits
    • Cloud credits to scale your IT product
    • Gateway to Russia
    • Cloud for Startups
    • Center for Technologies and Society
    • Yandex Cloud Partner program
    • Price calculator
    • Pricing plans
  • Customer Stories
  • Documentation
  • Blog
© 2026 Direct Cursus Technology L.L.C.
Yandex Managed Service for PostgreSQL
  • Getting started
    • Resource relationships
    • Planning a cluster topology
    • High availability clusters
    • Networking in Managed Service for PostgreSQL
    • Quotas and limits
    • Storage in Managed Service for PostgreSQL
    • Backups
    • Assigning roles
    • Managing connections
    • Replication
    • Maintenance
    • Supported clients
    • PostgreSQL settings
    • Indexes
    • SQL command limits
    • PostgreSQL versioning policy
  • Access management
  • Pricing policy
  • Terraform reference
  • Monitoring metrics
  • Audit Trails events
  • Public materials
  • Release notes

In this article:

  • PostgreSQL version lifecycle terms and stages
  • Features for each version
  • Key principles of the PostgreSQL versioning policy
  • Update policy
  • Automatic update and stopping procedure
  • Notifications
  • Versioning schedule
  1. Concepts
  2. PostgreSQL versioning policy

PostgreSQL versioning policy

Written by
Yandex Cloud
Updated at April 9, 2026
  • PostgreSQL version lifecycle terms and stages
    • Features for each version
  • Key principles of the PostgreSQL versioning policy
  • Update policy
  • Automatic update and stopping procedure
  • Notifications
  • Versioning schedule

This document describes the lifecycle of major PostgreSQL versions in Yandex Cloud. The policy is based on the PostgreSQL community's official support cycle and seeks to strike a balance between stable cluster operation and maintaining a modern, secure infrastructure.

Note

The new major version (New) becomes available shortly after its release by the PostgreSQL community. At the initial stage, not all functions and extensions may be fully supported. We recommend you to consult the guides for the current list of supported extensions.

PostgreSQL version lifecycle terms and stagesPostgreSQL version lifecycle terms and stages

In Managed Service for PostgreSQL, each major PostgreSQL version has the following 5-year lifecycle:

Version status Description and key actions Period1
New Latest version with long-term support (LTS). Recommended for all new projects. First year
Supported Previous LTS version. Fully supported; allows creating new clusters. Recommended for all projects. Years two through four
Deprecated Version nearing the end of its support period. The existing clusters operate normally. Six months before the end of support, an active notification effort starts about the need to update. From this point on, you cannot create new clusters. Fifth year
End of life (EOL) Discontinued version. Clusters still running this version get auto-updated to the latest supported version or stopped. At the end of five years

1 The periods are relative to the major version release date. The exact status change dates are officially announced by the service.

Features for each versionFeatures for each version

Depending on PostgreSQL version status, the following operations are available for clusters:

Action New Supported Deprecated End of life (EOL)
Creating new clusters
Recovery from a backup
Use of existing clusters
Choice of the major version update window N/A

Warning

Long-term backups (LTR) created for versions close to EOL may not be available for restoration in Managed Service for PostgreSQL after that version is no longer supported. We recommend you to plan data migration before the end of the version lifecycle.

Key principles of the PostgreSQL versioning policyKey principles of the PostgreSQL versioning policy

The PostgreSQL versioning policy is based on the following key principles:

  • Support duration. Each major version remains either New or Supported for four and a half years.
  • Incremental deprecation. In its fifth year, the version get the Deprecated status, and it becomes impossible to create new clusters six months before the end of life date.
  • Active notification: Six months before EOL, Managed Service for PostgreSQL begins an active campaign to notify customers about the need to update.
  • End of life: On reaching EOL (end of fifth year), clusters running obsolete versions get **automatically updated ** to the latest Supported version. If the update is not possible for technical reasons, the cluster will be stopped. Before it stops, the final backup will be automatically uploaded to the customer's Object Storage bucket. After the cluster is stopped and the backup is uploaded, the cluster will be deleted.

Update policyUpdate policy

  • Minor updates (within a major version, e.g., 15.1 → 15.2) are installed automatically as part of maintenance performed during the cluster's specified maintenance window or at the customer's request. These updates contain security and bug fixes. The update requires a short, successive reboot of the cluster hosts.
  • Major updates (major version changes, e.g., 15.x → 16.x) are user-initiated, except for automatic updates for versions that have reached EOL. We highly recommend you to update before the EOL date.

Automatic update and stopping procedureAutomatic update and stopping procedure

For clusters still running versions with expired support (EOL), Managed Service for PostgreSQL initiates an automatic update.

  1. Notification: Owners of clusters based on Deprecated versions will get a series of official email notifications: the first one 6 months before EOL, and then 90 and 30 days before the scheduled start date of the auto-update campaign.
  2. Automatic update: Three months before EOL, Managed Service for PostgreSQL begins the process of auto-updating clusters that had not been updated by their owners.
  3. Stopping as a final measure: If an automatic update is not possible, Managed Service for PostgreSQL will stop the cluster. Before it stops, the final backup will be created and saved in the customer's bucket.
  4. Access to data: Once the cluster is stopped, the responsibility for data is transferred to the customer. You can restore the data only in your own infrastructure, e.g., Yandex Compute Cloud VMs, using the backup you export.

NotificationsNotifications

Managed Service for PostgreSQL provides these upcoming update notifications in advance:

  • Entering Deprecated status (prohibits creating new clusters).
  • Six months ahead of EOL (start of active notification phase).
  • Start of the automatic update campaign: 90 and 30 days before the start of the campaign.

Warning

We recommend you to schedule cluster updates in advance before you get your auto-update notifications. This will allow you to choose a convenient time for migration and avoid the risks associated with automatic updates.

Versioning scheduleVersioning schedule

The current status of major PostgreSQL versions is based on the official support schedule.

Version2 New Supported Deprecated End of life (EOL)
PostgreSQL 18 2025–2026 2025–2029 2030 End of 2030
PostgreSQL 17 — 2024–2028 2029 End of 2029
PostgreSQL 16 — 2023–2027 2028 End of 2028
PostgreSQL 15 — 2022–2026 2027 End of 2027
PostgreSQL 14 — — 2026 End of 2026

2 Exact dates of transitions between statuses will be announced additionally. The EOL dates match the PostgreSQL community end of support dates.

Was the article helpful?

Previous
SQL command limits
Next
Access management
© 2026 Direct Cursus Technology L.L.C.