New and changed in version 26.3

New features and changes in version 26.3

SQL editor enhancements

The SQL editor is now enhanced with these features:

  • Simplified database object management - Easily create and manage database objects, including schemas, tables, views, stored procedures, and flex tables through an intuitive interface.
  • Real-time SQL preview - Instantly view the generated SQL at every step with greater transparency and control before applying changes.
  • Enhanced object explorer - Quickly browse and navigate all database objects from a centralized object explorer.
  • Detailed object insights - Select any database object to view its properties, configuration details, and relevant metadata in a dedicated information panel.
  • Improved SQL editor experience - Execute multiple SQL queries in a single session for quicker and efficient query workflows.
  • Advanced query analysis tools - Analyze query behavior with integrated support for viewing execution plans and explain plans, helping users optimize performance and troubleshoot queries effectively.

Additional supported platforms

Unified Console now supports all platforms supported by Management Console (MC), including x86 and ARM64 architectures, giving you greater flexibility to deploy and manage Vertica across a wider range of hardware environments.

Query optimization enhancements

Enhanced query optimization capabilities in Vertica AI Assist empower users with:

  • LLM-powered explanations: Summarize query optimizations and provides clear, actionable insights for troubleshooting and performance tuning.
  • Metadata-aware optimization: Leverage comprehensive Vertica metadata, including table statistics, projections, and resource usage, to deliver more accurate and context-aware optimization rationale.
  • Visual explain plan graphs: Render optimized explain plans in a graphical interface, allowing intuitive analysis of query execution paths, operators, and costs.
  • Profile summary: Provide a high-level overview of query execution, including duration, status, and resource usage.
  • Query profile export: Generate profile export details and a download link for the profile tarball (.tar.gz).

Vertica Spark Connector 4.1.1 with Java 21 support

You can now use Vertica Spark Connector 4.1.1 in Java 21 environments. This update includes a fix for cross-database JDBC cache reuse, improving the reliability of Data Source V2 (DSv2) read and write operations.

Vertica Grafana Connector supports Grafana 13

You can query, explore, and visualize data with full support for Grafana 13 dashboards. This provides compatibility with Grafana 13 and includes the updates required to align with Grafana's React 19 frontend framework and plugin architecture.

Management Console supports multi-factor authentication (MFA)

Administrators can configure MFA for local authentication, federated server authentication, and identity provider (IDP) authentication workflows through Management Console's integration with Keycloak. Supported IDP protocols include SAML v2.0, OpenID Connect v1.0, and Keycloak OpenID Connect.

Object-level in-database restore points

You can now save object-level in-database restore points by specifying object inclusion rules when you create an archive or save a restore point. Object-level restore points include only the specified objects and their dependent objects, and you can restore all or a subset of those objects from the archive. For more information, see Save restore points and RESTORE FROM ARCHIVE.

NUMA multi-node cluster management with VCluster CLI

VCluster CLI now supports managing NUMA (Non-Uniform Memory Access) multi-node clusters, where multiple database nodes run on the same physical host on different client ports. New options for add_node and remove_node let you specify per-node client ports and control NUMA node assignment for each host. This works both on real NUMA hardware and in standard environments running multiple nodes on a single host. For more information, see NUMA multi-node clusters.

Standalone MCP server RPM

Starting in version 26.3, the Vertica MCP server is distributed as a standalone RPM with its own release cadence, and must be installed on a dedicated host outside the Vertica cluster. Going forward, each MCP server release supports the most recently released version of the Vertica database. Note that MCP server release 26.3 specifically requires Vertica version 26.2.0-1, as this patch has been updated with standalone MCP server support.

For installation instructions, see Installing the MCP server RPM.

TLS v1.3 support

Vertica now supports TLS v1.3 in addition to TLS v1.2. The following changes provide support for both protocol versions:

  • CREATE TLS CONFIGURATION and ALTER TLS CONFIGURATION now include a separate CIPHERLIST parameter for specifying TLS v1.2 cipher suites. The previously existing CIPHER SUITES parameter is now used for TLS v1.3 cipher suites exclusively.

  • The TLS_CONFIGURATIONS system table includes a new CIPHER_LIST column for TLS v1.2 cipher suites.

  • A new TLSCipherSuites Security parameters specifies TLS v1.3 cipher suites for client-server communication. The existing EnabledCipherSuites parameter continues to apply to TLS v1.2.

Run end-to-end ML workflows with the MCP server

The Vertica MCP server now provides machine learning (ML) tools for running end-to-end, in-database ML workflows through natural-language interaction with AI agents. Using conversational prompts, you can prepare data, split, train and test sets, train and evaluate models, generate predictions, and manage models across their lifecycle.

The tools are organized into categories such as data preparation and feature engineering, exploratory data analysis, model training, prediction, model evaluation, and model registry and governance. For an overview and example workflows, see Machine learning with the MCP server. For a complete reference of every tool, see ML tools reference.

TLS fallthrough annotation for VerticaDB operator deployments

The VerticaDB operator supports a new vertica.com/enable-fallthrough annotation that controls whether operator-generated TLS authentications allow fallthrough. When set to true, a failed TLS connection can fall through to password authentication. This release also adds detailed guidance for enabling TLS across the database lifecycle on Kubernetes when creating, reviving, or updating a database, as well as during operator and server upgrades. For more information, see Custom resource definition parameters and Enable TLS in Vertica on Kubernetes.

Expose additional ports on subcluster services

The VerticaDB custom resource now supports a subclusters[i].extraPorts parameter that lets you expose additional ports on a subcluster service. Each entry accepts a port name, port, optional targetPort, and nodePort (for NodePort services), so you can route traffic for services such as metrics or administration endpoints alongside the standard database ports. For more information, see Custom resource definition parameters.

Export additional log streams with the sidecar logger

The vertica-logger sidecar image now supports exporting logs from different database components. Beginning with vertica-logger version 2.1.0, you can set the LOG_FILE_TYPE environment variable to export a specific log stream. Set it to VERTICA (default) for vertica.log, SPREAD for spread.log, or UDXLOGS for UDxFencedProcesses.log. Each sidecar exports a single log stream, so add a separate sidecar for each component you want to export. For more information, see VerticaDB custom resource definition.

ARM64 architecture support for containerized images

Vertica container images now support the arm64 architecture in addition to amd64. Any image with multiarch in its tag runs on both amd64 and arm64 hosts. For an example, see the opentext/vertica-logger tags. For more information, see Vertica images.