This Innovation Release includes features and enhancements.
Release Date: May 19, 2026
Release cadence update
HM now follows a dual release strategy: LTS and Innovation Releases. For details on this new model, see Hybrid Manager (HM) dual release strategy.
Release highlights
Expanded Postgres-to-Postgres schema migration: The Data Migration Service (DMS) now handles Postgres-specific database objects during schema migration, including custom domain types and other Postgres-native type definitions, enabling complete end-to-end schema migration between Postgres databases.
Enhanced agentic migration: The Migration Agent Langflow template now uses an improved architecture leveraging HM's unified schema migration feature, with added support for Llama 3 Nemotron 49B and Nano 4B models.
HM chatbot enhancements: Added Anthropic Claude as a chatbot provider and bundled HM and community PostgreSQL documentation for citation-backed answers.
Role mapping improvements: You can now assign multiple roles in a single role mapping rule, streamlining identity provider group-to-role assignments.
Release details
AI Factory
| Type | Description |
|---|---|
| Enhancement | Added support for AIDB 7 in the AI Pipeline Designer, including improved knowledge base validation and fixes for issues affecting pipelines with existing knowledge base tables and EDB Postgres Distributed (PGD) clusters. For full details, see the AIDB 7 release notes. |
| Enhancement | Added support for registering external inference services with insecure connections. You can now enable the Allow insecure connection option when registering external model services that don't have signed certificates, such as those in test environments. The system requires explicit user acknowledgment of the security risk before allowing the registration. |
| Enhancement | Enhanced error handling when publishing AI flows with invalid parameters. The UI now displays detailed error messages from the server instead of a generic Failed to publish flow message, helping you identify and fix parameter naming issues. |
| Change | Removed the AI Pipelines Default Settings page from the Sovereign AI section. The default settings feature was incompatible with the new instance-specific models list approach. |
| Bug fix | Resolved a startup failure in the EDBAirman component where uvx dependency resolution errors prevented the Model Context Protocol (MCP) server from initializing. The runtime now includes all required dependencies. |
HM console
| Type | Description |
|---|---|
| Enhancement | Added the beacon.enterprisedb.com/exclude-storage-from-dbaas=true label to exclude specific StorageClass or VolumeSnapshotClass resources from the HM console. In shared Kubernetes environments, you can apply this label to storage classes that aren't suitable for Postgres workloads—such as GPU- or HM-only classes—to prevent them from appearing in the Create Cluster dropdowns. The label takes effect on the next HM polling cycle and doesn't require a restart. |
| Enhancement | Enhanced the role mapping rules interface to support selecting multiple roles for a single rule. You can now assign multiple roles (such as both Project Viewer and Catalog Data Writer) to the same group and identity provider combination in a single rule, rather than creating separate rules for each role. |
| Enhancement | Added Anthropic Claude as a supported provider for the HM chatbot. The chatbot now renders Mermaid diagrams inline in responses. Bundled HM and community PostgreSQL (16/17/18) documentation with the chatbot so it can answer product questions with citations. |
Observability and monitoring
| Type | Description |
|---|---|
| Bug fix | Thanos Ruler used the default Kubernetes service account. It now uses a dedicated service account. |
Analytics
| Type | Description |
|---|---|
| Enhancement | Updated the PGAA version shipped with HM to 1.9. |
Migrations
| Type | Description |
|---|---|
| Enhancement | Added concurrent schema finalization for schema_and_data migrations. After the snapshot stage completes, you can select Finalize Schema Concurrently on the migration detail page to apply indexes, foreign keys, and other schema objects in the background using non-blocking data definition language (DDL) variants. When you later complete the migration, only lightweight validation steps remain, significantly reducing cutover time. |
| Enhancement | Extended Postgres-to-Postgres schema migration in the DMS to handle Postgres-specific database objects—including custom domain types and other Postgres-native type definitions—during schema migration. The DMS now ensures it applies these schema objects to the target before other schema objects that depend on them, enabling complete, end-to-end schema migration between Postgres databases. |
| Enhancement | Enhanced the Simple Migration Agent Langflow template with an improved architecture that leverages the unified schema migration feature, replacing the legacy Migration Portal approach. The updated workflow includes streamlined agent components, automatic password generation, and database setup, with redundant components removed. |
| Enhancement | Added support for Llama 3 Nemotron 49B and Llama 3 Nemotron Nano 4B models in the Simple Migration Agent Langflow template, providing additional AI model options for migration workflows. |
| Enhancement | Enhanced database password security in the Simple Migration Agent Langflow template by moving cluster creation operations outside of large language model (LLM) processing and leveraging global secrets management to prevent credential exposure. |
| Enhancement | Optimized the migration creation form to skip unnecessary table loading and validation when you set the migration scope to schema-only. The form now only fetches table data for data-inclusive migration scopes, improving performance and user experience. |
| Bug fix | Fixed an issue in the Simple Migration Agent Langflow template whereby the list of tools in an MCP Tools component wouldn't update after selecting an MCP server until you refreshed the page. |
| Bug fix | Added support for migrating EDB Postgres Advanced Server (EPAS) PACKAGE and PACKAGE BODY schema objects from EPAS source databases to EPAS destination databases. The DMS now properly handles package and package body statements during schema migration, including correct dependency ordering to ensure the DMS applies package bodies after the tables they reference. |
| Bug fix | Added support for migrating EPAS non-public SYNONYM schema objects from EPAS source projects. The DMS now properly handles non-public synonym statements during schema migration. The DMS extracts and uploads PUBLIC SYNONYM objects into the Migration Portal project but doesn't yet support migrating them to the destination database. |
| Bug fix | Fixed an issue whereby the schema migration capability couldn't migrate database objects for schemas that didn't contain any tables. The migration process now handles all schemas regardless of whether they include tables. |
| Bug fix | Fixed an issue whereby schema-only migrations dropped PRIMARY KEY constraints and their dependent foreign key and named CHECK constraints when you disabled data migration for specific schemas but enabled it for others. The transporter now correctly applies PRIMARY KEY constraints during the Finalize Schema phase for tables excluded from data migration. |
| Bug fix | Fixed an issue whereby the DMS validation incorrectly required all schemas in the source database to have a healthy Migration Portal project, even when the migration didn't include those schemas. The validation now only checks schemas that the migration's table list references. |
| Bug fix | Fixed an issue whereby the transformed SQL file had missing line endings when the source SQL file contained Windows-style CR LF line endings. |