- Introduced a high‑level RESTful API (sharing the gRPC port) with added endpoints for vector retrieval by ID, JSON ‘contains’ queries, bulk insert with partition keys, and GCS/OSS chunk manager support.
- Enhanced ANN search capabilities: vector fields can be returned, retrieval performance improved for HNSW/DiskANN/IVF FLAT, reduced overhead for large top‑K, and better write performance in partition‑key and multi‑partition scenarios.
- Resolved numerous bugs (excessive disk usage, compaction stalls, delete inconsistencies, MinIO/OSS failures, bulk insert crashes, memory leaks) and added monitoring, configuration, and CPU usage optimizations.