Skip to main content
Version: Early Access

Digital.ai Deploy 25.3.x Release Notes

Early Access (EA) Releases​

Deploy Operator​

Support Policy​

For each version of Deploy, we provide maintenance support for 15 months. For more information, see Digital.ai Support Policy.

Upgrade Instructions​

The Digital.ai Deploy upgrade process you use depends on the version from which you are upgrading, and the version to which you want to go.

For upgrade instructions, see:

Breaking Changes​

Breaking changes in Deploy 25.3 require your attention before upgrading:

  • Deploy no longer performs runtime checks for the presence of tar or unzip on the target host. These utilities must be pre-installed on deployment targets. If they are missing, archive-based copy strategies will fail.
    • For Linux: Ensure both tar and unzip are installed before deployment.
    • For Windows: Ensure PowerShell 5.0+ with unzip support is available.
  • When uploading a tar file, always set the copy strategy to Tar. For a zip file, use ZipWindows or ZipUnix as appropriate for the target OS.
  • The recommended approach for file.Folder deployments is to use the Tar strategy for optimal performance.

Updated System Requirements​

Deploy 25.3 supports the following latest versions of Java, operating systems, and databases. Before upgrading, ensure your environment meets the following requirements:

  • JDK: Supports OpenJDK 21
    (Oracle JDK 17 and OpenJDK 17 are no longer supported)

  • Windows Server: 2025, 2022
    (2019 is no longer supported)

  • RHEL: 9.x, 10.x
    (8.x is no longer supported)

  • PostgreSQL: 17.5, 16.9
    (17.2 and 16.6 are no longer supported)

  • MySQL: 8.4 LTS
    (8.0 LTS is no longer supported)

  • SQL Server: 2022 (16.0)
    (2019 (15.0) is no longer supported)

For more information, see Installation Prerequisites.

RabbitMQ 4.x and Quorum Queues​

From operator version 25.3 or later, RabbitMQ is upgraded to version 4.x, and quorum queues are now used by default instead of mirrored classic queues.

For more details, see Manual Upgrade RabbitMQ.

Provision Deploy Applications in Backstage Using YAML​

You can now provision Digital.ai Deploy applications in Backstage using either inline YAML or YAML files from GitHub. This feature allows you to define your Deploy application configuration as YAML and apply it directly to Digital.ai Deploy through Backstage, supporting both manual and GitOps-compatible workflows.

For more information, refer to Provisioning Deploy Applications in Backstage Using YAML.

API Versioning and Revisions in Azure API Management​

You can now manage API versions and revisions in Azure API Management directly through Digital.ai Deploy. This enhancement allows you to define API version sets, create multiple versions of an API, and manage non-breaking updates using revisions all within your deployment workflow. With this support, you can evolve APIs safely, maintain backward compatibility, and streamline versioned API deployments without manual steps in the Azure portal.

For more information, refer to Azure Plugin.

Personal Access Token (PAT) Support for Deploy​

You can now use Personal Access Tokens (PATs) for secure, passwordless authentication with Digital.ai Deploy APIs. PATs offer a safer alternative to username-password authentication, especially for automated use cases. Users can manage tokens through the new Access Tokens tab in their profile, with options to set custom expiration periods ranging from 7 days to no expiration. Administrators can enforce maximum token lifespans through system settings. A Expires in column provides clear visibility into token validity and expiration warnings. PAT support is also available for SSO-authenticated users, such as those using Office 365, Okta, or Azure AD. Organizations can control whether SSO users are allowed to generate PATs through a system-level setting, ensuring access remains aligned with the SSO provider’s active user and role state.

For more information, see Personal Access Token for Authentication.

Plugin Synchronization and Management Enhancements​

Digital.ai Deploy 25.3 introduces improved plugin handling across both the Plugin Manager UI and CLI. Core plugins such as base-plugin, command-plugin, and database-plugin are now visible in the Installed section of the UI and can be upgraded, deleted, or reinstalled directly. A dedicated plugin metadata file ensures consistent classification of official and custom plugins during installation. The CLI now supports official and local arguments to control installation paths, and in offline environments, plugins are installed based on these flags even without metadata access. Additionally, Deploy now removes duplicate plugin entries from the _local_ folder when an official version is installed, resolving upgrade conflicts.

For more information, refer to Plugin Synchronization.

Upload Deployment Artifacts to Google Cloud Storage​

You can now upload deployment artifacts including files, folders, and archives from your deployment package to a Google Cloud Storage (GCS) bucket. This feature supports both full replacement uploads and intelligent folder syncing, similar to AWS S3 functionality.

For more information, refer to Google Compute Plugin.