Skip to main content
Version: Next

1.45.0 - 2026-03-11

Added​

  • Flink: Add DatasetConfigFacet support for Flink native listener #4368 @jsingh-yelp Add support for emitting DatasetConfigFacet in the Flink native listener, enabling configuration tracking for datasets processed by Flink jobs.
  • Spec: Add HierarchyDatasetFacet #3747 @dolfinus Introduce the HierarchyDatasetFacet to provide structured representation of dataset hierarchy levels (database, schema, table, etc.) without relying on dataset name parsing, enabling consistent handling across different database systems with varying hierarchy depths.

Changed​

  • Spark: Replace semanticHash with identity-based VisitedNodes tracking #4383 @mobuchowski Improve performance by switching from expensive semanticHash() calls to identity-based tracking using IdentityHashMap, eliminating hot path in large jobs during plan traversal.
  • Spark: Use HashSet instead of LinkedList for BFS visited set #4376 @mobuchowski Optimize findDependentInputs by replacing LinkedList with HashSet for visited node tracking, improving lookup performance from O(n) to O(1).

Fixed​

  • dbt: Fix singular tests handling in structured logs #4372 @mobuchowski Fix handling of dbt singular tests when processing structured logs to ensure proper test result tracking and lineage extraction.
  • Spark: Fix BigLake catalog detection condition #4357 @tnazarew Fix the condition for adding project_id to catalog properties, ensuring proper BigLake catalog detection and handling.