{"full_name":"milvus-io/milvus","owner":"milvus-io","name":"milvus","url":"https://github.com/milvus-io/milvus","description":"Milvus is a high-performance, cloud-native vector database built for scalable vector ANN search","homepage":"https://milvus.io","topics":[],"language":"Go","license":"Apache-2.0","avatar":"https://avatars.githubusercontent.com/u/51735404?v=4","created_at":"2019-09-16T06:43:43Z","pushed_at":"2026-11-10T09:36:00Z","metrics":{"stars":46629,"forks":4325,"watchers":333,"issues":17417,"prs":33972,"releases":180,"issues_open":null,"issues_closed":null,"prs_open":null,"prs_closed":null,"prs_merged":null,"discussions":3117},"growth":{"day":{"stars":15,"forks":1,"base_day":"2026-11-09"}},"history":{"dates":["2026-11-09","2026-11-10"],"stars":[46614,46629],"forks":[4324,4325],"watchers":[333,333],"issues_open":[null,null],"prs_open":[null,null]},"boards":{},"same_owner":[],"issues":[{"number":23232,"title":"[Feature]: Support HNSW SQ","comments":42,"created_at":"2023-04-04T21:29:19Z"},{"number":31581,"title":"[Feature]: Will milvus suppoert BGE-M3['Colbert'] storage  and search? ","comments":37,"created_at":"2024-03-25T09:05:36Z"},{"number":39027,"title":"[Enhancement]: Merge insert message of multi partitions in one shard as one message.","comments":33,"created_at":"2025-01-07T01:39:21Z"},{"number":24272,"title":"[Feature]: Support Graceful Shutdown of Worker Nodes in Milvus","comments":23,"created_at":"2023-05-19T21:38:51Z"},{"number":25343,"title":"[Feature]: Distinct field values","comments":22,"created_at":"2023-07-05T17:25:27Z"},{"number":2771,"title":"Add support for ScaNN index","comments":21,"created_at":"2020-07-07T16:46:57Z"},{"number":41035,"title":"[Feature]: Support gRPC-based custom tokenizer","comments":20,"created_at":"2025-04-01T10:18:52Z"},{"number":27577,"title":"[Feature]: Support date && datetime type ","comments":20,"created_at":"2023-10-09T13:03:02Z"}],"discussions":[{"number":55078,"title":"Error while vendoring tantivy-binding dependencies to build Milvus","comments":1,"category":"Q&A and General discussion","created_at":"2026-11-03T19:10:57Z"},{"number":54554,"title":"Best Practices for Milvus Schema Migrations Across Environments","comments":1,"category":"Q&A and General discussion","created_at":"2026-10-22T04:00:19Z"},{"number":54507,"title":"Where can I download the official RPM/DEB packages? (Documentation links return 404)","comments":1,"category":"Q&A and General discussion","created_at":"2026-10-21T05:29:34Z"},{"number":54073,"title":"milvus集群在大文件的时候，dataNode执行索引构建任务速度慢","comments":2,"category":"Q&A and General discussion","created_at":"2026-10-09T10:35:23Z"},{"number":54048,"title":"Woodpecker: segments with V6 footer (LAC) can't be read after downgrading Milvus 3.0.x -> 2.6.x (failed to decode footer record: buffer too short for record: expected N, got 45) — is downgrade supported?","comments":1,"category":"Q&A and General discussion","created_at":"2026-10-08T14:53:59Z"},{"number":53969,"title":"Using Milvus for video highlights: time-ranged evidence rows instead of text chunks","comments":1,"category":"Show and tell","created_at":"2026-10-06T08:03:03Z"},{"number":53952,"title":"實驗","comments":1,"category":"Announcement","created_at":"2026-10-03T05:08:02Z"},{"number":53904,"title":"[Helm] Pulsar 4.0.x incompatible with milvus-helm 5.0.x — bookkeeper.conf and OTEL config not updated for 4.x class paths — cross-ref: milvus-helm#318","comments":7,"category":"Q&A and General discussion","created_at":"2026-09-29T05:47:43Z"}],"latest_release":{"name":"milvus-3.0.5","tag":"v3.0.5","published_at":"2026-10-30T10:43:28Z","body":"## v3.0.5\r\n\r\nRelease date: October 30, 2026\r\n\r\n| Milvus Version | Python SDK Version | Node.js SDK Version | Java SDK Version | Go SDK Version |\r\n| -------------- | ------------------ | ------------------- | ---------------- | -------------- |\r\n| 3.0.5          | 3.0.2              | 3.0.6               | 3.0.12           | 3.0.0          |\r\n\r\nWe are excited to announce the release of Milvus v3.0.5! This release adds resource-group-aware load status in QueryCoord, reduces memory use in several paths. It also fixes stability issues around coordinator restarts, rolling upgrades and transient etcd errors.\r\n\r\n### Improvements\r\n\r\n- Added per-resource-group load percentage and shard-leader readiness in QueryCoord, and tagged each shard leader with its replica's resource group, so queries can be routed within a single resource group ([#52500](https://github.com/milvus-io/milvus/pull/52500), [#52711](https://github.com/milvus-io/milvus/pull/52711))\r\n- Reduced copying in unscored Tantivy term queries by borrowing decoded postings blocks directly ([#54469](https://github.com/milvus-io/milvus/pull/54469))\r\n- Upgraded Knowhere to v3.0.17 and moved nullable vector offset mapping behind Knowhere APIs ([#54473](https://github.com/milvus-io/milvus/pull/54473))\r\n- Reduced memory allocations when encoding batched RESTful query and search responses by reusing per-request encoding buffers ([#54540](https://github.com/milvus-io/milvus/pull/54540))\r\n- Improved text analyzer performance by reusing compiled regexes and per-filter search caches in analyzer filters ([#54570](https://github.com/milvus-io/milvus/pull/54570))\r\n- Added an opt-in two-tier mix compaction planner (`dataCoord.compaction.twoTierCompaction`, default off) to reduce compaction write amplification ([#54589](https://github.com/milvus-io/milvus/pull/54589))\r\n- Reduced timestamp index memory usage when loading small sealed segments ([#54659](https://github.com/milvus-io/milvus/pull/54659))\r\n- Added the missing `status=\"fail\"` label to the `internal_storage_op_count` metric for failed `get` operations; the old unlabeled series stops and the new series starts from zero ([#54637](https://github.com/milvus-io/milvus/pull/54637))\r\n- Allowed hyphens in resource group names after the first character, matching the naming rules for partitions ([#54679](https://github.com/milvus-io/milvus/pull/54679))\r\n- Reduced memory allocations in compaction merge sort by reusing output columns across records ([#54718](https://github.com/milvus-io/milvus/pull/54718))\r\n\r\n### Bug fixes\r\n\r\n- Fixed an issue where collections with struct array fields could force-load raw embedding-list vectors on first search or after cache eviction, increasing memory usage and query latency ([#54167](https://github.com/milvus-io/milvus/pull/54167))\r\n- Fixed an issue where upserts on CMEK-enabled collections wrote the delete part (primary keys and timestamps) to the WAL without encryption ([#54478](https://github.com/milvus-io/milvus/pull/54478))\r\n- Fixed an issue where `ST_DWITHIN` filters returned fewer rows on segments with an RTREE index than on unindexed segments; empty geometries no longer match `ST_DWITHIN`, consistent with PostGIS ([#54611](https://github.com/milvus-io/milvus/pull/54611))\r\n- Fixed an issue where writes to Storage V3 segments could stall after a coordinator restart because the next flush used a column layout that conflicted with the committed manifest ([#54652](https://github.com/milvus-io/milvus/pull/54652))\r\n- Fixed an issue where QueryNode could panic while replaying encrypted WAL messages from a dropped database that shared a physical channel with other collections ([#54640](https://github.com/milvus-io/milvus/pull/54640))\r\n- Fixed an issue where partial upserts failed on collections with nullable struct array fields when the field was omitted and the batch contained new primary keys ([#54619](https://github.com/milvus-io/milvus/pull/54619))\r\n- Fixed an issue where index state was incorrectly reported "},"readme":true,"readme_zh":true,"news":[],"stale":false,"snapshot_at":"2026-11-10T09:35:25Z","history_start":"2026-09-06"}