Release 478-e STS (8 Dec 2025)#

Starburst Enterprise platform (SEP) 478-e is the follow up SEP release to 478-e. It includes all improvements from the following Trino releases:

This release is a short term support (STS) release.

Breaking changes#

As of SEP 478-e, the default implementation of the ConnectorFactory#getSecuritySensitivePropertyName method is abstract. This method identifies configuration properties that may contain security-sensitive information so the engine can mask their values. Custom connectors must provide their own implementation. This change affects only customers who maintain custom plugins.

Upgraded all SEP versions to run on Java 25.0.1.
Added support for configuring log levels in the Starburst Data Catalog using the starburst-portal Helm chart.
Added automatic sanitization of database names in Starburst data catalog when defaultLocationUri is set. Database names are sanitized before being appended to the S3 location URI to ensure compliance with AWS Glue requirements for valid characters and maximum URI length.
Removed active catalog properties from query.json in troubleshooting output.
Fixed an issue where viewing query history failed when traffic was routed through Starburst Gateway configured with an Oracle backend.

Added support for public OAuth endpoints to ensure compatibility with GCP token passthrough.

Improved handling of inline tool calls for AI Agent. It now detects cases where OpenAI or Bedrock SDKs return tool calls inline within the text response and automatically prompts the model to correct the response format and retries the request.
Improved handling of long-running Agent sessions by introducing automated session compaction.
Improved Agent SQL generation by incorporating SQL language documentation.

Added support for multi-field pushdown.
Fixed an issue where wildcard queries failed when targeting indices with inconsistent field mappings.

Added support for writing partition statistics.
Added support for setting and dropping default column values.
Fixed an issue that caused incorrect results when reading Avro files migrated from Hive.
Fixed data inconsistency and silent data loss issues under heavy write load when using the Iceberg connector with Starburst data catalog.

Added the mongodb.schema-database configuration property, which specifies where schema information for MongoDB collections is stored.

Improved raw_query performance by adding the opensearch.sharded-scrollable-query-passthrough-enabled flag, which lets each shard returns its results as a single JSON payload.