v3.0.2

## v3.0.2

Release date: September 20, 2026

| Milvus Version | Python SDK Version | Node.js SDK Version | Java SDK Version | Go SDK Version |
| -------------- | ------------------ | ------------------- | ---------------- | -------------- |
| 3.0.2          | 3.0.2              | 3.0.6               | 3.0.10           | 3.0.2          |

We are excited to announce the release of Milvus v3.0.2! This release focuses on search and query performance — removing hot-path contention in filtered search, group-by, and index building — along with stronger external collection and Storage V2 support, and a broad set of stability fixes across streaming, compaction, and index management.

### Improvements

- Optimized ARRAY filtering by merging chained contains predicates into a single `ContainsAny`/`ContainsAll` expression in the query planner ([#52365](https://github.com/milvus-io/milvus/pull/52365))
- Added support for custom S3-compatible endpoints in external collections via the `extfs.endpoint_url` option, with validation of unsafe or conflicting endpoint settings ([#52814](https://github.com/milvus-io/milvus/pull/52814))
- Unified Bloom and Roaring membership filters behind a single `membership_match` expression, with matching Go client support for building and querying both filter types ([#53019](https://github.com/milvus-io/milvus/pull/53019))
- Reduced write amplification during Tantivy-based index building, lowering disk I/O for text match, NGRAM, and JSON key stats indexes ([#53057](https://github.com/milvus-io/milvus/pull/53057))
- Added admission control on RESTful v2 DQL endpoints that returns HTTP 429 with `Retry-After` before request decoding when the proxy query queue is full ([#53111](https://github.com/milvus-io/milvus/pull/53111))
- Reduced an atomic refcount hotspot in the scalar filter evaluation path that accounted for ~48% of leaf CPU time in search, improving filtered search throughput ([#53167](https://github.com/milvus-io/milvus/pull/53167))
- Improved storage thread pool scalability by replacing the hand-rolled implementation with `folly::CPUThreadPoolExecutor` and restoring elastic worker scaling ([#53184](https://github.com/milvus-io/milvus/pull/53184))
- Improved REST access log matching so formatters are matched against the parsed URL path, and configured methods now apply to requests carrying query parameters ([#53147](https://github.com/milvus-io/milvus/pull/53147))
- Added idempotent broadcast support so retried requests no longer create duplicate tasks, with `BulkImport` as the first adopter ([#53228](https://github.com/milvus-io/milvus/pull/53228))
- Added configurable sentence split characters for the Lindera tokenizer, allowing user dictionary entries containing punctuation to match as a single token ([#53287](https://github.com/milvus-io/milvus/pull/53287))
- Added a cluster-version gate that automatically enables write-before function materialization only after all nodes finish upgrading, avoiding mixed-version inconsistency during rolling upgrades ([#53261](https://github.com/milvus-io/milvus/pull/53261))
- Upgraded Woodpecker to v0.1.42, fixing WAL recovery failures on finalized empty segments and improving append-path stability and metrics ([#53295](https://github.com/milvus-io/milvus/pull/53295))
- Reduced per-chunk atomic and reference-count overhead in the search and query hot path by pinning a sealed segment snapshot once per request ([#53301](https://github.com/milvus-io/milvus/pull/53301))
- Improved partial update latency by replacing TimeTick waits with snapshot-based optimistic locking, and refined AutoID handling so existing primary keys are preserved and returned IDs keep the input order ([#53337](https://github.com/milvus-io/milvus/pull/53337))
- Reduced redundant row-view construction when grouping search results by VARCHAR or JSON fields on sealed segments, lowering group-by search overhead ([#53500](https://github.com/milvus-io/milvus/pull/53500))
- Added a compliance API that reports load configuration convergence globally and per resource group, covering replica serviceability, query visibility, residual resources and WAL placement ([#53517](https://github.com/milvus-io/milvus/pull/53517))

### Bug fixes

- Fixed an issue where queries could return incorrect results after a struct array field was dropped and re-added ([#52921](https://github.com/milvus-io/milvus/pull/52921))
- Fixed SASL/SCRAM-SHA-256 authentication failures when connecting to Apache Kafka 4.x brokers by upgrading librdkafka to 2.6.1 ([#53086](https://github.com/milvus-io/milvus/pull/53086))
- Fixed an issue where all channels of a replica were assigned to a single query node, causing repeated out-of-memory kills and leaving the replica unserviceable ([#53094](https://github.com/milvus-io/milvus/pull/53094))
- Fixed repeated WAL fencing that stalled writes on a PChannel for 45-60 seconds every few minutes under sustained ingest ([#53118](https://github.com/milvus-io/milvus/pull/53118))
- Fixed Storage V2 compaction dropping valid physical group paths, which could misalign packed column indexes in compacted segments ([#53202](https://github.com/milvus-io/milvus/pull/53202))
- Fixed compaction logs printing collection encryption keys and object storage credentials ([#53226](https://github.com/milvus-io/milvus/pull/53226))
- Fixed stale struct array offset aliases that could surface incorrect data after a sealed segment was reopened ([#53154](https://github.com/milvus-io/milvus/pull/53154))
- Fixed file resource synchronization running before any resource was added, which could clear node-local files on startup or node registration ([#53170](https://github.com/milvus-io/milvus/pull/53170))
- Fixed an issue where an incomplete binlog chunk could be silently treated as fully read, risking missing data in query and compaction results ([#53263](https://github.com/milvus-io/milvus/pull/53263))
- Fixed an issue where storage of dropped segments was never reclaimed for collections without any active index ([#53252](https://github.com/milvus-io/milvus/pull/53252))
- Fixed inaccurate resource estimation when loading sparse vector indexes, which could lead to incorrect raw-data handling and repeated warnings on QueryNode ([#53249](https://github.com/milvus-io/milvus/pull/53249))
- Fixed an issue where a frozen streaming node outside the primary resource group could be unfrozen unexpectedly during rebalancing ([#53229](https://github.com/milvus-io/milvus/pull/53229))
- Fixed a failure to connect to Google Cloud Storage where GCP credentials (IAM and HMAC) were not registered before the chunk manager pre-check ([#53288](https://github.com/milvus-io/milvus/pull/53288))
- Fixed an issue where C++ logs were unexpectedly written to the `/tmp` directory instead of being forwarded to the unified log output ([#53293](https://github.com/milvus-io/milvus/pull/53293))
- Fixed backfill commits failing with an HTTP 500 error when a Spark result spanned multiple partitions ([#53346](https://github.com/milvus-io/milvus/pull/53346))
- Fixed a memory leak in DataNode where a canceled index-build or analyze task never released the native memory of the object it had already built ([#53348](https://github.com/milvus-io/milvus/pull/53348))
- Fixed an issue where binlog import failed when a nullable vector field had no binlog files ([#53363](https://github.com/milvus-io/milvus/pull/53363))
- Fixed incomplete or incorrect output fields when search and query requests read external table data ([#53372](https://github.com/milvus-io/milvus/pull/53372), [#53385](https://github.com/milvus-io/milvus/pull/53385))
- Fixed duplicated sparse-vector rows and partition IDs in REST requests, plus memory ownership and cleanup issues that could cause crashes or leaks when queries exited early ([#53402](https://github.com/milvus-io/milvus/pull/53402))
- Fixed an issue where a compaction reporting completion without a result payload could crash DataCoord or leave the compaction task stuck instead of being cleanly retried ([#53443](https://github.com/milvus-io/milvus/pull/53443))
- Fixed an issue where properties of external data files were dropped when building segment manifests, causing external collections to lose source file metadata ([#53444](https://github.com/milvus-io/milvus/pull/53444))
- Fixed a QueryNode crash that could occur when handling `count(*)` requests at debug log level, notably during rolling upgrades ([#53474](https://github.com/milvus-io/milvus/pull/53474))
- Fixed an issue where index and stats build tasks kept consuming worker resources after their segment, index, or collection had been dropped, and improved cleanup of orphaned index files ([#53515](https://github.com/milvus-io/milvus/pull/53515))
- Fixed local storage path handling so that data written by different components always lands where readers and garbage collection expect it, with automatic upgrade for existing local deployments ([#53530](https://github.com/milvus-io/milvus/pull/53530))
- Fixed index build tasks retrying forever when a segment contained malformed JSON documents; such builds now fail fast instead of consuming worker resources indefinitely ([#53531](https://github.com/milvus-io/milvus/pull/53531))
- Fixed concurrent snapshot restores targeting the same collection: restores are now serialized and a restore to an existing target is rejected with a clear "already exists in database" error instead of racing or leaking resources ([#53586](https://github.com/milvus-io/milvus/pull/53586))
- Fixed binary encryption keys being corrupted across storage interfaces and misaligned cursors in projected Storage V2 packed readers by upgrading milvus-storage ([#53569](https://github.com/milvus-io/milvus/pull/53569))