1 Release published by 1 person

• v2.5.13 milvus-2.5.13

  published Jun 9, 2025

75 Pull requests merged by 21 people

• feat: Add json flat index

  #39917 merged Jun 10, 2025

• enhance: remove extra get collection

  #42042 merged Jun 10, 2025

• enhance: refactor rrf and weighted rerank

  #42154 merged Jun 10, 2025

• fix: Use delete checkpoint to prevent delete record loss in L0 refactoring

  #42628 merged Jun 10, 2025

• enhance: Adjust the default parameters of WP according to performance tests

  #42598 merged Jun 10, 2025

• fix: Release data memory after sync task completes

  #42627 merged Jun 10, 2025

• fix: avoid load index or create interim index in ChunkedSegmentSealedImpl::HasRawData()

  #42622 merged Jun 10, 2025

• fix: timetick cannot push forward when upgrading

  #42567 merged Jun 10, 2025

• enhance: milvus support minhash vector and mhjaccard metric

  #42036 merged Jun 10, 2025

• fix: Fix the bug of valid data write corruption

  #42556 merged Jun 10, 2025

• fix:[2.5] add concurrency and close protect for bm25 function

  #42599 merged Jun 10, 2025

• fix: Add explicit move semantics to get_batch_view interface

  #42403 merged Jun 10, 2025

• enhance: [2.5] bulkinsert handles nullable/default

  #42072 merged Jun 10, 2025

• fix: [StorageV2] sync with int8 vector data type core dumped

  #42616 merged Jun 10, 2025

• test: use service dns name as tei endpoint

  #42565 merged Jun 10, 2025

• fix: add concurrency and close protect for bm25 function

  #42597 merged Jun 10, 2025

• enhance: [2.5] Set thread name for segcore thread pool

  #42596 merged Jun 10, 2025

• fix: [StorageV2] index/stats task level storage v2 fs

  #42191 merged Jun 10, 2025

• enhance: Upgrade pymilvus to 2.5.11rc7

  #42458 merged Jun 10, 2025

• enhance: use english name as language name for all type language identifier

  #42600 merged Jun 10, 2025

• enhance: [StorageV2] Pass bucket name for compaction readers

  #42607 merged Jun 10, 2025

• enhance: [storageV2] Pass nullable converting insertMsg fieldData

  #42584 merged Jun 10, 2025

• fix: filter the streaming query node from resource group when upgrading

  #42594 merged Jun 9, 2025

• enhance: support auto index type for json index

  #42071 merged Jun 9, 2025

• enhance: support auto index type for json index

  #42161 merged Jun 9, 2025

• fix: ReleasePartition cause delegator unserviceable.

  #42423 merged Jun 9, 2025

• [automated] Bump milvus version to v2.5.13

  #42610 merged Jun 9, 2025

• fix: Check cast type is array for json contains expr

  #42184 merged Jun 9, 2025

• [automated] Bump milvus version to v2.5.13

  #42605 merged Jun 9, 2025

• [automated] Bump milvus version to v2.5.13

  #42606 merged Jun 9, 2025

• enhance: [storagev2] Support partition key isolation index

  #42574 merged Jun 9, 2025

• enhance: Adjust default import buffer size

  #42541 merged Jun 9, 2025

• test:Add TTL read test

  #42189 merged Jun 9, 2025

• fix: storage v2 write mmap file per field per cell

  #42180 merged Jun 9, 2025

• fix: update tantivy commit version to fix stemmer panic

  #42171 merged Jun 9, 2025

• fix issue where wp fast-fail pending appends leads to client retry duplicated write

  #42566 merged Jun 9, 2025

• feat: implement batch commit for JSON Stats

  #42494 merged Jun 8, 2025

• fix: pipeline/delegator leak

  #42583 merged Jun 6, 2025

• fix: pipeline/delegator leak

  #42582 merged Jun 6, 2025

• fix: Fix delegator selection logic in releaseSegment

  #42572 merged Jun 6, 2025

• fix: [2.5] Fix the bug of valid data write corruption

  #42555 merged Jun 6, 2025

• fix: lost metrics of flowgraph and write buffer

  #42549 merged Jun 6, 2025

• enhance: Increase search/query retry times on proxy before timeout

  #40438 merged Jun 6, 2025

• enhance: Make segcore thread name set correctly

  #42497 merged Jun 6, 2025

• fix: Check cast type is array for json contains expr

  #42185 merged Jun 6, 2025

• enhance: [2.5] update maintainers

  #42558 merged Jun 5, 2025

• enhance: Support cast function for json index

  #42504 merged Jun 5, 2025

• enhance: Support cast function for json index

  #41949 merged Jun 5, 2025

• enhance: [2.5] Adjust default import buffer size

  #42542 merged Jun 5, 2025

• enhance: [2.5] Accelerate dispatcher building

  #42544 merged Jun 5, 2025

• enhance: update maintainers

  #42539 merged Jun 5, 2025

• fix: Fix intermittent deadlock in grouped allocator

  #42524 merged Jun 5, 2025

• fix: fix is null bug for marisa index

  #42420 merged Jun 5, 2025

• enhance: Accelerate dispatcher building

  #42500 merged Jun 5, 2025

• fix: querynode upgrade from 2.5 get stucked

  #42503 merged Jun 5, 2025

• fix: add ShowLoadCollections and ShowLoadPartitions for compatibale mixcoord

  #42514 merged Jun 5, 2025

• fix: Just trigger stats task for Flushed segment

  #42424 merged Jun 5, 2025

• enhance: run analyzer should get database name from grpc context

  #42398 merged Jun 5, 2025

• fix: analyzer memory leak because function runner not close

  #41839 merged Jun 5, 2025

• fix: Pass cluster id tranforming drop task to drop job request

  #42531 merged Jun 5, 2025

• fix: streamingnode get stucked when stop

  #42501 merged Jun 5, 2025

• feat: limit search result entries

  #42522 merged Jun 5, 2025

• test: Fix unstable integration test for partial result

  #42511 merged Jun 5, 2025

• fix stuck insert by binding buffer and chan to ensure proper notification

  #42505 merged Jun 5, 2025

• fix: [2.5] Fix duplicate autoID between import and insert

  #42520 merged Jun 4, 2025

• fix: Only mark segment compacting for sort stats task

  #42516 merged Jun 4, 2025

• fix: [2.5] Only mark segment compacting for sort stats task

  #42517 merged Jun 4, 2025

• fix: Fix duplicate autoID between import and insert

  #42519 merged Jun 4, 2025

• enhance: Add proxy task queue metrics

  #42156 merged Jun 4, 2025

• enhance: refine compaction trigger to reduce read/write amplifaction(#41336)

  #41728 merged Jun 4, 2025

• fix: querynode upgrade from 2.5 get stucked

  #42502 merged Jun 4, 2025

• feat: [Tiered] Make load list work as warmup hint

  #42490 merged Jun 4, 2025

• fix: release memory after pop from heap

  #42482 merged Jun 4, 2025

• enhance: avoid shutdown sdk api in minio cm destructor

  #42459 merged Jun 4, 2025

• fix: Fix sort stats generates large binlogs

  #42456 merged Jun 4, 2025

33 Pull requests opened by 19 people

• enhance: Skip rate limit for superusers

  #42508 opened Jun 4, 2025

• fix: Fix l0 segment leak by refactoring segment checker logic

  #42525 opened Jun 4, 2025

• enhance: [2.5] Restrict partition key when requery

  #42528 opened Jun 4, 2025

• enhance: support json stats with shredding design

  #42534 opened Jun 5, 2025

• enhance: [2.5] Skip rate limit for superusers

  #42543 opened Jun 5, 2025

• enhance: [2.5] support print nq and params for search and query

  #42545 opened Jun 5, 2025

• enhance: add test for new parameters in access log

  #42546 opened Jun 5, 2025

• enhance: Move sort stats task to sort compaction

  #42562 opened Jun 5, 2025

• fix: Fix exclude nodes clearing logic position in load balancer retry

  #42577 opened Jun 6, 2025

• enhance:enable zstd compress by default (#42585)

  #42586 opened Jun 7, 2025

• enhance: allow '.' and '-' characters in usernames (#42417)

  #42588 opened Jun 7, 2025

• test: add default user and pwd

  #42590 opened Jun 9, 2025

• test: [cp2.5]add default user and pwd

  #42591 opened Jun 9, 2025

• enhance: Optimize memory usage during garbage collection

  #42593 opened Jun 9, 2025

• enhance: [2.5] use english name as language name for all type language identifier

  #42601 opened Jun 9, 2025

• test: update sdk version and cases[2.5]

  #42603 opened Jun 9, 2025

• test: update ttl test comments and update for expressions tests

  #42611 opened Jun 9, 2025

• enhance: Enhance import integration tests and logs

  #42612 opened Jun 9, 2025

• fix: Unsorted small segments should not be considered as indexed

  #42614 opened Jun 9, 2025

• fix: [2.5] Unsorted small segments should not be considered as indexed

  #42615 opened Jun 9, 2025

• feat: add back queryNode.cache.warmup for compatibility

  #42621 opened Jun 9, 2025

• build(deps): bump requests from 2.26.0 to 2.32.4 in /tests/python_client

  #42625 opened Jun 10, 2025

• fix: limit the concurency of zstd compression and decrease the memory usage of binlog generation

  #42630 opened Jun 10, 2025

• enhance: [2.5] Optimize memory usage during garbage collection

  #42631 opened Jun 10, 2025

• fix: [AddField] Use corresponding datatype for int8/int16 def val

  #42633 opened Jun 10, 2025

• enhance: test pymilvus correct(#38436)

  #42637 opened Jun 10, 2025

• enhance: [StorageV2] sync separate vector datatype into its own column group

  #42638 opened Jun 10, 2025

• enhance: Check loaded segments before gc

  #42639 opened Jun 10, 2025

• enhance: Support run analyzer by collection field

  #42642 opened Jun 10, 2025

• fix: Model rerank supports Truncate

  #42643 opened Jun 10, 2025

• fix: compatibility with old sessions upgrade from 2.5 to 2.6 in standalone mode

  #42645 opened Jun 10, 2025

• enhance: add segment bm25 stats local cache (#41775)

  #42646 opened Jun 10, 2025

• enhance: Replace pointer-based map key with id in garbage collector

  #42647 opened Jun 10, 2025

49 Issues closed by 11 people

• [Bug]: [dn&in] The dataNode was oomkilled while bulk_import multi parquet files

  #42608 closed Jun 10, 2025

• [Bug]: The order of results obtained from Milvus search + model reranker is inconsistent with the order of results obtained from Milvus search + manually calling the API reranker.

  #42619 closed Jun 10, 2025

• [Bug]: [benchmark][cluster] when caching layer warmup.vectorIndex=disable, the queryNode load the index files again after pod failure is recovered

  #42526 closed Jun 10, 2025

• [Bug]: [streaming] The streamingNode's memory usage increased by 2GB during quiescence unexpectedly

  #40627 closed Jun 10, 2025

• [Bug]: The mixCoord's memory usage increased by 1GB+ during quiescence unexpectedly

  #41121 closed Jun 10, 2025

• [Bug]: [woodpecker] Concurrent inserts are stuck, leading to growing CP lag

  #42049 closed Jun 10, 2025

• [Bug]: TextLog Didn't respect mmap enabled

  #42535 closed Jun 10, 2025

• [Bug]: Milvus panic when running e2e tests

  #42617 closed Jun 10, 2025

• [Bug]: [streaming] When the length of varchar or json of some entities is long, the memory usage of queryNodes is very uneven

  #40971 closed Jun 10, 2025

• [Bug]: [streaming] Standalone memory usage unexpectedly increased by almost 5 GiB during quiescence

  #41202 closed Jun 10, 2025

• [Bug]: shardDelegator leak

  #42581 closed Jun 10, 2025

• [Bug]: The streaming node consumes a large amount of memory, making it prone to OOM (Out of Memory).

  #40540 closed Jun 9, 2025

• [Bug]: [IvfRabitq] Describe index does not return the exact value as users input when creating index

  #41783 closed Jun 9, 2025

• [Bug]: test: limit resouce for ci and nightly tests

  #41819 closed Jun 9, 2025

• [Bug]: [json_path_index] Create json path index on non-json field does not report error

  #40636 closed Jun 9, 2025

• [Bug]: The version for 2.5.5 image is "pkg/v2.5.5" rather than "v2.5.5"

  #40216 closed Jun 9, 2025

• [Bug]: [Nightly] Load occasionally failed reporting "wait for loading collection timeout"

  #39844 closed Jun 9, 2025

• [Bug]: [null & default] Group by field does not support null

  #36264 closed Jun 9, 2025

• [Bug]: In the standalone deployment mode, under scenarios of concurrent reading and writing, a large number of read requests fail.

  #41655 closed Jun 9, 2025

• [Bug]: [rerank] The distance difference between the score calculated by the formula and the returned distance by search becomes larger as the increase of the field value used for rerank

  #41509 closed Jun 9, 2025

• [Bug]: Many operations like load, release and flush failed after milvus recovered from many chaos test due to mixcoord panic when release segment

  #42568 closed Jun 9, 2025

• [Bug]: flowgraph metrics lost after streaming is enabled

  #42548 closed Jun 9, 2025

• [Bug]: tantivy merge may cause doc id out of order

  #41597 closed Jun 7, 2025

• [Bug]: CVE in Milvus 2.5.6

  #41641 closed Jun 7, 2025

• [Bug]: [dn&in] The dataNode was oom killed in concurrent DML & DQL & scene_search_test scene

  #41333 closed Jun 6, 2025

• [Enhancement]: Defer the unlink operation until after munmap

  #41636 closed Jun 6, 2025

• [Bug]: [benchmark][standalone] It took 1 hour to build the FLAT index, because the flushing segments are used for the stats task

  #42419 closed Jun 6, 2025

• [Bug]: [rerank] Search results are less than expected with rerank after release, drop index, recreate index and load

  #42011 closed Jun 6, 2025

• [Bug]: [streaming] Standalone terminated with log SIGNO: 11; SIGNAME: Segmentation fault

  #41704 closed Jun 6, 2025

• [Bug]: [streaming] The partition has been dropped but its related segments are still there

  #41976 closed Jun 6, 2025

• [Bug]: [Nightly] Query mix string expression without indexes are missing results

  #41086 closed Jun 6, 2025

• [Bug]: The search result is not the same with the first one after multiple partition release and collection reload

  #42038 closed Jun 6, 2025

• [Bug]: [benchmark][cluster] hybrid search raises error `fail to Query on QueryNode 2: worker(4) query failed: segment not found: segment not loaded[segment=455484340280804555]` in concurrent ddl & dql scene

  #39540 closed Jun 6, 2025

• [Bug]: [benchmark][cluster][master] DQL requests raise error query failed: segment not found: segment not loaded in dql & dml scene(enabled partition key)

  #41833 closed Jun 6, 2025

• [Bug]: [benchmark][cluster] DQL requests raise error `query failed: segment not found: segment not loaded` in dql & dml scene(enabled partition key)

  #38893 closed Jun 6, 2025

• [Bug]: [Nightly] Many cases failed for load collection timeout (Milvus mode is cluster)

  #41654 closed Jun 6, 2025

• [Bug]: Master use more cpu than 2.5 on growing segment search, the RT is greater than 2.5 too.

  #42039 closed Jun 6, 2025

• [Bug]: The memory for loading tens of millions of sets has been stuck for over 2 hours

  #41414 closed Jun 6, 2025

• [Enhancement]: Tidy all serialize and deserialzie before read/write to object storage

  #39242 closed Jun 6, 2025

• [Bug]: [Add Field]When creating indexes for multiple added fields, it will cause the execution of the compaction task to fail

  #42022 closed Jun 5, 2025

• [Bug]: [Add Field] Parse output field name failed: $meta["extra_info"]

  #41702 closed Jun 5, 2025

• [Enhancement]:BM25 Grouping Search Indexing

  #41944 closed Jun 5, 2025

• [Bug]: [benchmark][standalone] search raises error `fail to search on QueryNode 1: channel not found` in concurrent DML & multi-partition scene

  #42174 closed Jun 5, 2025

• [Bug]: milvus-standalone unhealthy

  #41538 closed Jun 5, 2025

• [Bug]: Milvus still use many Go Memory after recovering is done

  #42481 closed Jun 5, 2025

• [Bug]: [benchmark][cluster] dataNode(16G) OOM during building indexes

  #42400 closed Jun 4, 2025

• [Enhancement]: restful describe index api should return index params

  #42015 closed Jun 4, 2025

• [Bug]: [rerank] There is no rules for the name of the Function during rerank

  #41530 closed Jun 4, 2025

• [Enhancement]: Refine Drop Partition

  #41542 closed Jun 4, 2025

45 Issues opened by 26 people

• [Feature]: Supprt null vectors

  #42651 opened Jun 10, 2025

• [Bug]: `pymilvus 2.5.11` with `milvusdb/milvus:v2.5.13-gpu` : Incorrect port or sdk is incompatible with server

  #42650 opened Jun 10, 2025

• [Bug]: WAL cannot be recovered if create operation of pulsar producer fail

  #42649 opened Jun 10, 2025

• [Enhancement]: If the distances in the ANN search results are the same, entities with larger pk values should rank higher, rather than the current situation where smaller pk values rank higher.

  #42644 opened Jun 10, 2025

• [Bug]: [Add Field]during the concurrent search process, creating index for new field will cause a queryNode exception

  #42640 opened Jun 10, 2025

• [Bug]: no result from client search

  #42636 opened Jun 10, 2025

• [Bug]: [benchmark][standalone] Milvus standalone OOM when flushing a collection

  #42634 opened Jun 10, 2025

• [Bug]: Modal rerank function lacks truncate related params

  #42632 opened Jun 10, 2025

• [Bug]: [Add Field] append null/default values to arrow builder failed

  #42629 opened Jun 10, 2025

• [Feature]: Text Match is messy

  #42626 opened Jun 10, 2025

• [Bug]: Modulo operator in filter does not work correctly with JSON data

  #42624 opened Jun 9, 2025

• [Bug]: error msg does not indicate which field's index params are wrong if there are multiple fields creating index

  #42620 opened Jun 9, 2025

• [Bug]: [StorageV2] sync with int8 vector data type core dumped

  #42613 opened Jun 9, 2025

• [Bug]: querying with expression 'varchar != '1'' returns None if TRIE index built

  #42604 opened Jun 9, 2025

• [Bug]: [mixcoord] When upgrading version 2.5 to master, the new standalone cannot running: old session exists

  #42602 opened Jun 9, 2025

• [Enhancement]: Adjust the default parameters of WP according to performance tests

  #42595 opened Jun 9, 2025

• [Bug]: MixCoord OOMs during garbage collection of dropped segments

  #42592 opened Jun 9, 2025

• [Feature]: Support highlight

  #42589 opened Jun 8, 2025

• [Enhancement]: Enable zstd compress by default

  #42585 opened Jun 6, 2025

• [Bug]: After the Linux server was unexpectedly shut down, I encountered the following error when restarting the Milvus Docker standalone instance:

  #42580 opened Jun 6, 2025

• [Bug]: .query taking keyword args is very risky

  #42579 opened Jun 6, 2025

• [Bug]: tokenizer_c report heap-use-after-free when e2e test

  #42576 opened Jun 6, 2025

• [Bug]: Model rerank function design is not clean in my opinion

  #42573 opened Jun 6, 2025

• [Bug]: [woodpecker] After inserting 20 millio entities serially, both count(*) and binlog show that there are more data than expected

  #42563 opened Jun 5, 2025

• [Bug]: shard leader cache doesn't refresh during retry

  #42561 opened Jun 5, 2025

• [Enhancement]: Move the sort stats task into the sort compaction

  #42560 opened Jun 5, 2025

• [Bug]: The query node seems to be returning offset+limit rows instead of trimming the offset

  #42559 opened Jun 5, 2025

• [Bug]: Scalar index null data will be corrupted due to multiple calls to FillFieldData

  #42554 opened Jun 5, 2025

• [Bug]: Full-text search crashes Milvus service when using Java SDK on Chinese Windows.

  #42553 opened Jun 5, 2025

• [Bug]: Model rerank function does not support setting api key

  #42552 opened Jun 5, 2025

• [Bug]: Model rerank function does not support vllm version 0.7.x

  #42551 opened Jun 5, 2025

• [Bug]: WebUI always describes collections in the default database

  #42550 opened Jun 5, 2025

• [Bug]: Milvus query node does not work when using AWS S3 with pod-identity.

  #42547 opened Jun 5, 2025

• [Bug]: Milvus server crashes when processing search query with filter '(f0 / (2877 / -4571)) <= -3731'

  #42538 opened Jun 5, 2025

• [Bug]: gosdk 2.5 milvusclient.Client.Search() cannot handle Vector(float16) correctly

  #42537 opened Jun 5, 2025

• [Enhancement]: support json stats with shredding methods

  #42533 opened Jun 5, 2025

• [Bug]: CI asnyc testcase failed with error `AttributeError: 'AsyncGrpcHandler' object has no attribute 'deregister_state_change_callbacks'`

  #42532 opened Jun 5, 2025

• [Bug]: Drop task fails when task execution failed on datanode worker

  #42530 opened Jun 5, 2025

• [Bug]: Assertion Failure: Assert "(value_proto.val_case() == milvus::proto::plan::GenericValue::kInt64Val)"

  #42527 opened Jun 4, 2025

• [Bug]: Fix intermittent deadlock in grouped allocator

  #42523 opened Jun 4, 2025

• [Feature]: Limit search result entries to protect milvus from OOM error

  #42521 opened Jun 4, 2025

• [Bug]: Duplicate autoID between import and insert

  #42518 opened Jun 4, 2025

• [Bug]: [woodpecker] Standalone panic: send on closed channel during serial insertion

  #42513 opened Jun 4, 2025

• [Bug]: Unstable integration test for partial result

  #42510 opened Jun 4, 2025

• [Bug]: After upgrading from 2.5 to master, two sealed-L1 segments were left unsorted and more data than expected

  #42506 opened Jun 4, 2025

91 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• feat: impl StructArray -- data storage support in segcore

  #42406 commented on Jun 10, 2025 • 39 new comments

• feat: supporing load priority for loading

  #42413 commented on Jun 6, 2025 • 6 new comments

• [Bug]: [benchmark][standalone] milvus suddenly OOM in concurrent DDL scene

  #42129 commented on Jun 10, 2025 • 0 new comments

• [Bug]: [Add Field] Sometimes flush time after delete with concurrent add field operation is larger than that without concurrent add field operation

  #41871 commented on Jun 10, 2025 • 0 new comments

• [Bug]: [2.5.x] [etcd] etcdserver: invalid auth token

  #39445 commented on Jun 10, 2025 • 0 new comments

• [Bug]: [streaming] After upgrading from 2.5+ to master, the all dql and dml requests failed: DEADLINE_EXCEEDED

  #42492 commented on Jun 10, 2025 • 0 new comments

• [Enhancement]: [streaming] Standalone cpu usage is high during concurrent inserts

  #41747 commented on Jun 10, 2025 • 0 new comments

• [Bug]: [Add Field]When the base database data is relatively large, adding multiple fields will cause streamingnode to crash

  #41941 commented on Jun 10, 2025 • 0 new comments

• [Bug]: [streaming] After upgrading from 2.5.11+ to master, the number of growing segments doubled

  #42162 commented on Jun 10, 2025 • 0 new comments

• [Bug]: [benchmark][standalone] milvus suddenly OOM when loading a collection in streaming mode

  #42081 commented on Jun 10, 2025 • 0 new comments

• [Bug]: sync task of flow graph should not panic if failed after enabling streaming service

  #39732 commented on Jun 9, 2025 • 0 new comments

• [Bug]: [Go-SDK] The SDK version requirement is too high and does not match the documentation

  #41748 commented on Jun 9, 2025 • 0 new comments

• [Bug]: Milvus continues to restart after a single instance writes tens of millions of data successfully

  #41601 commented on Jun 9, 2025 • 0 new comments

• [Bug]: [streaming] After upgrading from 2.5.11+ to master, wal assign failed and streamingNode dataNode were oomkilled

  #41616 commented on Jun 9, 2025 • 0 new comments

• [Bug]: [benchmark][cluster] load collection raises error `collection not loaded` in concurrent ddl(replica=2) & dql scene

  #39519 commented on Jun 9, 2025 • 0 new comments

• [Bug]: [woodpecker] Woodpecker memory usage is too high or even oom during concurrent inserts

  #41846 commented on Jun 9, 2025 • 0 new comments

• [Bug]:[woodpecker] In the DDL scene, when MQ is Woodpecker, the memory usage of SN is much higher than when MQ is Pulsar

  #41986 commented on Jun 9, 2025 • 0 new comments

• [Bug]: [json_path_index] The queried results after json path index created is less than that without index using "!=" or "not in" expression

  #40629 commented on Jun 9, 2025 • 0 new comments

• [Bug]: The error message is not user friendly when query with the expression including array with different data types

  #40644 commented on Jun 9, 2025 • 0 new comments

• [Bug]: The matching rule is not consistent when filtering with expression when there is "None" data for dynamic and non-dynamic field

  #40645 commented on Jun 9, 2025 • 0 new comments

• [Bug]: The "None" in the dynamic field could not be filtered using "field_name IS NULL" expression

  #40646 commented on Jun 9, 2025 • 0 new comments

• [Bug]: The queried results under expression “my_json['a'][0] == 1 && my_json['a'][1] == 2” is less than expected in large data amount scenario with no scalar index

  #40685 commented on Jun 9, 2025 • 0 new comments

• [Bug]: [json_path_index] Milvus is crashed and create index timeout after 2370 json path indexes created

  #40711 commented on Jun 9, 2025 • 0 new comments

• [Enhancement]: Remove MilvusStreamingServiceEnabled env in 2.6.x

  #41609 commented on Jun 6, 2025 • 0 new comments

• feat: impl StructArray -- support create collection with StructArrayField

  #42499 commented on Jun 4, 2025 • 0 new comments

• enhance: Make json index compatible with caching layer

  #42484 commented on Jun 10, 2025 • 0 new comments

• test:add test cases for add field

  #42472 commented on Jun 10, 2025 • 0 new comments

• feat: impl StructArray -- data storage support in go

  #42453 commented on Jun 4, 2025 • 0 new comments

• fix: Pre-check import message to prevent pipeline block indefinitely

  #42415 commented on Jun 10, 2025 • 0 new comments

• enhance: Add bulk api for json data

  #42407 commented on Jun 10, 2025 • 0 new comments

• fix: add describeIndex timestamp for restful interface

  #42104 commented on Jun 10, 2025 • 0 new comments

• test: add `add_collection_field` feature in checker for chaos test

  #42085 commented on Jun 6, 2025 • 0 new comments

• enhance: Remove unlink and use mmap chunk manager

  #41903 commented on Jun 9, 2025 • 0 new comments

• feat: optimize `Like` query with n-gram

  #41803 commented on Jun 4, 2025 • 0 new comments

• enhance: fix legacy mergify config

  #41693 commented on Jun 6, 2025 • 0 new comments

• just for test: Go core dump test (#41173)

  #41178 commented on Jun 10, 2025 • 0 new comments

• feat: Add CMEK encryption plugin

  #40323 commented on Jun 9, 2025 • 0 new comments

• [Bug]: After making frequent periodic writes to milvus-standalone, the milvus-standalone container becomes unhealthy

  #42065 commented on Jun 10, 2025 • 0 new comments

• [Bug]: Milvus (v2.5.12 in standalone mode) does not start, the milvus-standalone container terminates with the errors context deadline exceeded, fail to list policy, and status 134. Errors indicate problems connecting to etcd.

  #42479 commented on Jun 10, 2025 • 0 new comments

• [Enhancement]: Merge insert message of multi partitions in one shard as one message.

  #39027 commented on Jun 10, 2025 • 0 new comments

• [Bug]: [streaming] Query and search failed: channel distribution is not serviceable, required load ratio is 1.000000, current load ratio is 0.999025

  #42404 commented on Jun 10, 2025 • 0 new comments

• [Enhancement]: milvus super user can skip rate limit

  #42397 commented on Jun 10, 2025 • 0 new comments

• [Bug]: build dispatcher costs 2s which leads to the first search latency much higher

  #42067 commented on Jun 10, 2025 • 0 new comments

• [Bug]: [streaming] Query count(*) results in more than expected during chaos kill streamingNode container

  #41570 commented on Jun 10, 2025 • 0 new comments

• [Bug]: [benchmark][cluster][warmup=disable] DQL request raise error `fail to Query on QueryNode 3: channel not available[channel=channel distribution is not serviceable]` in concurrent DQL & multi-partition scene

  #42098 commented on Jun 10, 2025 • 0 new comments

• [Bug]: [benchmark][standalone] In the scenario of multiple growing segments, milvus continues to upsert and then OOM

  #42028 commented on Jun 10, 2025 • 0 new comments

• [Enhancement]: Remove the datanode's dependency on mixcoord

  #41611 commented on Jun 6, 2025 • 0 new comments

• [Bug]: SchemaError(io.milvus.v1alpha1.MilvusCluster.spec.components.dataCoord.volumes): array should have exactly one sub-item

  #32065 commented on Jun 6, 2025 • 0 new comments

• Create multiple milvus cluster on one dependency cluster

  #41672 commented on Jun 6, 2025 • 0 new comments

• [Enhancement]: pymilvus windows support

  #41635 commented on Jun 6, 2025 • 0 new comments

• [Bug]: Milvus docker image deleted after reboot

  #42478 commented on Jun 5, 2025 • 0 new comments

• [Bug]: [Add Field]Milvus Panic with error `panic: field new_field__YzgrbHtO not found when converting insert msg to insert data: field not found[field=116]` when Insert the corresponding data immediately after adding the field

  #42084 commented on Jun 5, 2025 • 0 new comments

• [Feature]: storage v2 support

  #39173 commented on Jun 5, 2025 • 0 new comments

• [Enhancement]: Search and query operation of proxy and delegator should retry until context deadline exceed

  #39828 commented on Jun 5, 2025 • 0 new comments

• [Bug]: [benchmark][cluster] queryNode sparse panic in concurrent dql & ddl scene

  #38295 commented on Jun 5, 2025 • 0 new comments

• [Enhancement]: Do logging at server-side when grpc exception with some common server information

  #38862 commented on Jun 5, 2025 • 0 new comments

• [Feature]:

  #26418 commented on Jun 4, 2025 • 0 new comments

• [Bug]: When deploying Milvus, authentication was enabled, but the web UI can still be accessed without a username/password.

  #40267 commented on Jun 4, 2025 • 0 new comments

• [Bug]: [streaming] Failed to connect to milvus for a long time after removing the proxy container-kill chaos

  #41331 commented on Jun 4, 2025 • 0 new comments

• [Feature]: Make partial load field list work as hint

  #42489 commented on Jun 4, 2025 • 0 new comments

• [Bug]: insert data error

  #41988 commented on Jun 4, 2025 • 0 new comments

• [Feature]: Faster filtering by IDs

  #30339 commented on Jun 4, 2025 • 0 new comments

• [Bug]: milvus can never start

  #41703 commented on Jun 4, 2025 • 0 new comments

• [Bug]: The Docker image won't start

  #42396 commented on Jun 4, 2025 • 0 new comments

• [Bug]: hybird search return error:Unexpected error: [hybrid_search], required argument is not a floa

  #42496 commented on Jun 4, 2025 • 0 new comments

• [Bug]: [benchmark][cluster] the CPU of queryNode is too low during concurrent search, causing search timeout 60s and failure

  #35728 commented on Jun 4, 2025 • 0 new comments

• [Bug]: boost::uuids::entropy_error

  #42062 commented on Jun 4, 2025 • 0 new comments

• [Bug]: [benchmark][cluster] flush raises error `failed to call flush to data coordinator: channel not found` in concurrent ddl & dql scene

  #39588 commented on Jun 4, 2025 • 0 new comments

• [Bug]: The error message is not user friendly when query with array with not same type like "json_contains_all(float_array, [2, 4.0])"

  #41176 commented on Jun 9, 2025 • 0 new comments

• [Bug]: [rerank] There is typo in the error message for the invalid value of offset during rerank Function

  #41533 commented on Jun 9, 2025 • 0 new comments

• [Bug]: [benchmark][standalone] query count* slow down to > 60s, the max RT of normal query count* on v2.5 does not exceed 10s

  #42163 commented on Jun 9, 2025 • 0 new comments

• [Bug]: [streaming] Search and hybrid_search failed: N6milvus21ExecOperatorExceptionE :Operator::GetOutput failed...Error in GetObjectSize[errcode:404

  #42412 commented on Jun 9, 2025 • 0 new comments

• [Bug]: In read-write concurrent scenarios for standalone, there are many query/search/hybrid search failures, and numerous slow queries.

  #41277 commented on Jun 9, 2025 • 0 new comments

• [Bug]: [streaming] During rolling upgrade image from master to master, some search failed: channel not available

  #42192 commented on Jun 9, 2025 • 0 new comments

• [Enhancement]: Allocate L0 Segment before append operation of wal to make L0 generation stable

  #42416 commented on Jun 9, 2025 • 0 new comments

• [Bug]: [benchmark][cluster] create collection raises error `failed to watch channels, code: NotReadyServe, reason: service not ready[mixture=1]: Initializing` after the instance is deployed successfully

  #42151 commented on Jun 9, 2025 • 0 new comments

• [Bug]: When there are duplicate values in the query, the required slop for phrase matching does not match the expected behavior

  #41738 commented on Jun 9, 2025 • 0 new comments

• [Enhancement]: Add memory protection on proxy

  #41692 commented on Jun 9, 2025 • 0 new comments

• [Enhancement]: New Search Architecture Based On Streaming Service

  #40451 commented on Jun 9, 2025 • 0 new comments

• [Enhancement]: Change the implementation of load operation into AsyncCGO mode

  #39138 commented on Jun 9, 2025 • 0 new comments

• [Enhancement]: Support Search With View Support And Milvus Streaming Service.

  #38125 commented on Jun 9, 2025 • 0 new comments

• [Bug]: error:SchemaError(io.milvus.v1alpha1.MitvusCtuster.spec.components.queryNode. initContainers):array should have exactly one sub-item

  #41668 commented on Jun 8, 2025 • 0 new comments

• [Bug]: Encountering raft::cuda_error for milvus index nodes on p4d.24xlarge AWS instance with 570.86.10 GPU driver

  #41633 commented on Jun 7, 2025 • 0 new comments

• [Bug]: [benchmark][cluster] queryNode loaded sorted_by_pk is false primary key raw data

  #42143 commented on Jun 6, 2025 • 0 new comments

• [Bug]: MessageID roll back when calling Woodpecker append operation

  #41842 commented on Jun 6, 2025 • 0 new comments

• [Bug]: [streaming] When concurrently upserting data, streamingNode is oomkilled and queries keep failing

  #40932 commented on Jun 6, 2025 • 0 new comments

• [Bug]: [streaming] Few search failed during pod-kill: worke query failed: segment not found: segment not loaded

  #41143 commented on Jun 6, 2025 • 0 new comments

• [Bug]: [streaming] It takes 30 minutes for StreamingNode to terminated and exit

  #41445 commented on Jun 6, 2025 • 0 new comments

• [Bug]: After upgrading from master to master, the old streamingNode cannot exit

  #42498 commented on Jun 6, 2025 • 0 new comments

• [Bug]: [Nightly][woodpecker] Many cases failed due to flush 180s timeout

  #41605 commented on Jun 6, 2025 • 0 new comments

• [Enhancement]: Support more ddl apis to replicate according to cdc

  #41677 commented on Jun 6, 2025 • 0 new comments