- Added APIs for modifying data stream backing indices, multivalue doc count aggregation, and Hive pull‑based ingestion.
- Introduced HTTP/3 client, gRPC support for PIT and bulk extra fields, and cross‑cluster streaming via Arrow Flight, plus performance and security enhancements.
- Fixed multiple bugs (timeout handling, NPEs, snapshot issues) and upgraded Jackson, Hadoop, Lucene, Netty and other libs to address CVEs.