Milvus

Milvus v2.3.16 focuses on stability, performance, and security. It includes bug fixes for auto-balancing, checkpoint handling, memory leaks, and transaction limits. Additionally, authentication for the REST v2 API has been strengthened, and query result retrieval has been optimized. Overall, this update aims to provide a more reliable, efficient, and secure platform for your needs.

Improvements

- Enhanced authentication for the REST v2 API.([33254](https://github.com/milvus-io/milvus/pull/33254))
- Improved checkpoint handling for message failures.([33159](https://github.com/milvus-io/milvus/pull/33159))([#33128](https://github.com/milvus-io/milvus/pull/33128))
- Improved the triggering mechanism for segment sync tasks so that it will only be triggered once by flushts ([33034](https://github.com/milvus-io/milvus/pull/33034))
- Improved the query result retrieval mechanism to get enough query results ([33177](https://github.com/milvus-io/milvus/pull/33177))

Bug fixes

- Fixed a bug that when suspending auto balance querynode will unexpectedly exit stopping balance ([32941](https://github.com/milvus-io/milvus/pull/32941))
- Resolved an issue leading to endless binlog path downloads during compaction ([33246](https://github.com/milvus-io/milvus/pull/33246))
- Addressed memory leaks in rendezvousflushmanager ([33112](https://github.com/milvus-io/milvus/pull/33112))
- Prevented system panics related to nil array field data ([33119](https://github.com/milvus-io/milvus/pull/33119))
- Fixed "SegmentNotLoaded" errors triggered by concurrent segment tasks in search ([33087](https://github.com/milvus-io/milvus/pull/33087))
- Fixed a bug that Etcd txn exceeds limit due to too many fields ([33049](https://github.com/milvus-io/milvus/pull/33049))
- Fixed a bug related to the incorrect calculation of the milvus_rootcoord_indexed_entity_num metric ([33002](https://github.com/milvus-io/milvus/pull/33002))

This release comes with several enhancements and bug fixes to improve security and stability. One of the improvements is the inclusion of the channel checkpoint in the flush response. Another enhancement is the ability to check whether autoID is enabled during data insertion, which optimizes the data migration using Milvus-CDC.

The bug fixes address issues such as empty segments in meta, and incorrect data types during data insertion according to the schema. A critical patch has also been implemented to solve the consistent failure and retries of the syncDistribution request to the delegator. We highly recommend updating to the latest version to benefit from these fixes.

Improvements

- Supported return channel checkpoint info in flush response ([32683](https://github.com/milvus-io/milvus/pull/32683))
- Add config to check id during inserting data when autoID enabled ([32840](https://github.com/milvus-io/milvus/pull/32840))

Critical Bug Fixes

- Fixing a bug where the syncDistribution request to the delegator consistently fails and retries persistently. ([32923](https://github.com/milvus-io/milvus/pull/32923))

Other Bug Fixes

- Fixed a bug where the system could generate empty segments. Added support for the compactor to clean up empty segments. ([32690](https://github.com/milvus-io/milvus/pull/32690))
- Fixed a bug that involved validating whether the input data types match the schema. ([32790](https://github.com/milvus-io/milvus/pull/32790), [#32845](https://github.com/milvus-io/milvus/pull/32845))

This release focuses on improving system performance, reliability, and observability with key features. It supports RESTful APIs to execute rolling upgrades, implements task-driven collection observation, allows for configurable intervals for garbage collection (GC), and enhances the check logic.

Furthermore, it fixes bugs in RESTful v2, resolving issues such as parameter parsing errors and ineffective parameter handling. Critical bug fixes related to deadlocks and data accuracy have also been included. Additionally, performance improvements have been made when handling large numbers of segments.

Overall, these updates aim to provide a more secure, efficient, and reliable user experience.

Improvements

- Added configuration items to skip Auto ID and Partition Key checks to improve check speed.([32671](https://github.com/milvus-io/milvus/pull/32671))
- Implemented the collection observer of QueryCoordV2 task-driven.([32615](https://github.com/milvus-io/milvus/pull/32615))
- Supported the get sdk type by user agent in access log.([32554](https://github.com/milvus-io/milvus/pull/32554))
- Supported the use of different intervals for GC scan.([32551](https://github.com/milvus-io/milvus/pull/32551))
- Removed the support for the always-true expression in delete expr.([32495](https://github.com/milvus-io/milvus/pull/32495))
- Allowed users to disable search optimization.([32143](https://github.com/milvus-io/milvus/pull/32143))
- Added restful api for devops to execute rolling upgrade.([31846](https://github.com/milvus-io/milvus/pull/31846))

Bug fixes

- Fixed the bug that the calculation precision may be different between growing index and sealed index.([32584](https://github.com/milvus-io/milvus/pull/32584))
- Fixed the bug that grow task stuck when node become stopping.([32556](https://github.com/milvus-io/milvus/pull/32556))
- Fixed the issue that leader view may not update when segment version did not match.([32517](https://github.com/milvus-io/milvus/pull/32517))
- Fixed the bug that value in elementTypeParams not be integer in restful api.([32362](https://github.com/milvus-io/milvus/pull/32362))
- Fixed dbName parameter of describeAlias request not take effect and fixed the issue which will report lack of permission when not pass the whole privileges to describe collection in restful api. ([32162](https://github.com/milvus-io/milvus/pull/32162))
- Fixed cp lag metrics leakage by marking channel checkpoint dropped.([32454](https://github.com/milvus-io/milvus/pull/32454))
- Fixed the bug where not update segment's version in syncDistribution.([32320](https://github.com/milvus-io/milvus/pull/32320))
- Fixed a bug that coordinator register blocked on ProcessActiveStandby.([32133](https://github.com/milvus-io/milvus/pull/32133))
- Fixed a bug that the reduce didn't handle offset without limit and reduceStopForBest correctly.([32087](https://github.com/milvus-io/milvus/pull/32087))
- Fixed a bug that not validata PlaceholderGroups before combine them.([32045](https://github.com/milvus-io/milvus/pull/32045))
- Fixed a bug where acquire index meta's lock for each segment.([31798](https://github.com/milvus-io/milvus/pull/31798))
- Fixed a bug that Rootcoord's stop may block in quota_center's stop.([31824](https://github.com/milvus-io/milvus/pull/31824))

This release aims to enhance system performance, reliability and observability with key features like improved index information retrieval, secure Minio connections via TLS, and new metrics for better monitoring.

This release has enriched the RESTful interfaces, allowing users to perform operations related to Partition and Index. Additionally, this release includes some critical bug fixes related to deadlocks and data integrity. Check the [RESTful API reference](https://milvus.io/api-reference/restful/v2.3.x/About.md) for more details.

These updates aim to provide a more secure, efficient, and reliable user experience.

Improvements

- Optimized DescribeIndex performance by implementing bulk index information retrieval([31239](https://github.com/milvus-io/milvus/pull/31239))([#31429](https://github.com/milvus-io/milvus/pull/31429))
- Improved access efficiency to formatted keys by caching the formatted key of param item([31402](https://github.com/milvus-io/milvus/pull/31402))
- Add metrics for querycoord current target cp lag([31463](https://github.com/milvus-io/milvus/pull/31463))
- Implemented TLS support for Minio connections([31292](https://github.com/milvus-io/milvus/pull/31292))([#31619](https://github.com/milvus-io/milvus/pull/31619))
- Speed up target recovery after QueryCoord restart([31449](https://github.com/milvus-io/milvus/pull/31449))
- Added new metrics for entities statistics([31511](https://github.com/milvus-io/milvus/pull/31511))
- Implementing the API of the resultful([30430](https://github.com/milvus-io/milvus/pull/30430))
- Save collection targets by batches([31655](https://github.com/milvus-io/milvus/pull/31655))
- Added validation checks for the legality of field data types([31699](https://github.com/milvus-io/milvus/pull/31699))

Bug Fixes

- Fixed the issue in CurrentTargetFirst/NextTargetFirst where it would generate unexpected task behavior([31419](https://github.com/milvus-io/milvus/pull/31419))
- Fixed the bug that does not cause the shard leader to fail when search fails([31450](https://github.com/milvus-io/milvus/pull/31450))
- Fixed a bug where nodes were marked as unreachable when get client failure([31451](https://github.com/milvus-io/milvus/pull/31451))
- Fixed client error handling to enable retries for certain types of unrecoverable errors instead of directly returning them([31452](https://github.com/milvus-io/milvus/pull/31452))
- Fixed an issue where the balance channel would freeze due to deadlocks([31455](https://github.com/milvus-io/milvus/pull/31455))
- Fixed a bug where nodeID was not checked when updating channel checkpoints([31508](https://github.com/milvus-io/milvus/pull/31508))
- Fixed the incorrect use of double buffering for entry size exceeding the max size([31549](https://github.com/milvus-io/milvus/pull/31549))
- Fixed a bug where delegator's filtering removed all delete messages, making deleted messages temporarily accessible([31587](https://github.com/milvus-io/milvus/pull/31587))
- Fixed a bug where the proxy's context was canceled prematurely during session revocation([31595](https://github.com/milvus-io/milvus/pull/31595))
- Fixed the target observer updates manual and automatic tasks were not mutually exclusive([31603](https://github.com/milvus-io/milvus/pull/31603))
- Fixed a bug where channel metadata was updated even when nodeID did not match([31665](https://github.com/milvus-io/milvus/pull/31665))
- Fixed a bug where compaction tasks were not cleaned up when releasing channels([31694](https://github.com/milvus-io/milvus/pull/31694))
- Fixed a bug where the injection was invalid when the queue was empty([31819](https://github.com/milvus-io/milvus/pull/31819))
- Fixed a bug where channel meta mergeFlushSegment was not idempotent cause data loss([31837](https://github.com/milvus-io/milvus/pull/31837))
- Fixed the bug that Datacoord session disconnects before revoking it([31747](https://github.com/milvus-io/milvus/pull/31747))

It is highly recommended to upgrade to this latest version to prevent data loss.

Bug Fixes

- [Critical] Fixed a bug that some insert data would be lost when deleteBuf memory policy is triggered ([31159](https://github.com/milvus-io/milvus/pull/31159))
- Fixed a bug that some duplicated segments couldn't be released after channel balancing ([31126](https://github.com/milvus-io/milvus/pull/31126))
- Fixed a bug that the balance checker of the Querycoord used the wrong param item for check interval ([31141](https://github.com/milvus-io/milvus/pull/31141))
- Fixed a bug that database info could be incorrect when describing collection with ID ([31177](https://github.com/milvus-io/milvus/pull/31177))

Improvements

- Change pagesize to 64k for AARCH64 platform ([31114](https://github.com/milvus-io/milvus/pull/31114))
- Introduce an internal API to list indexes meta and reduce memory usage during load collection ([31150](https://github.com/milvus-io/milvus/pull/31150))([#31163](https://github.com/milvus-io/milvus/pull/31163))

Milvus v2.3.11 brings improvements and bug fixes aimed at enhancing performance, security, and stability. The improvements include optimized data loading, enhanced security with TLS support for Kafka connections, memory optimization, and more. Additionally, bug fixes address issues such as search/query failures, incorrect data types in outputs, and disk estimation errors. We encourage you to update to this latest version to take advantage of these enhancements and fixes.

Improvements

- Optimized JSON loading by reducing 1x memory copy ([30864](https://github.com/milvus-io/milvus/pull/30864))
- Implemented TLS support for Kafka connections ([30466](https://github.com/milvus-io/milvus/pull/30466) [#30925](https://github.com/milvus-io/milvus/pull/30925))
- Optimized monitoring: Remove time tick delay metrics when nodes go offline ([30879](https://github.com/milvus-io/milvus/pull/30879))
- Optimized memory usage and loading speed for variable length data ([30900](https://github.com/milvus-io/milvus/pull/30900))
- Implemented support for varchar autoID in bulk insert operations ([30913](https://github.com/milvus-io/milvus/pull/30913))
- Added support for rate limiting for flush operations at the collection level ([29568](https://github.com/milvus-io/milvus/pull/29568))
- Improved compatibility for the watch DM channel request ([30954](https://github.com/milvus-io/milvus/pull/30954))
- Enhanced stability when reading S3 objects ([30976](https://github.com/milvus-io/milvus/pull/30976))
- Reduced contention for locks in the DescribeIndex operation ([30975](https://github.com/milvus-io/milvus/pull/30975))
- Optimized automatic balancing in QueryCoord ([30725](https://github.com/milvus-io/milvus/pull/30725))
- Enabled Milvus containers to run as a non-root user ([30937](https://github.com/milvus-io/milvus/pull/30937))
- Enhanced lock granularity in DataCoord meta ([30986](https://github.com/milvus-io/milvus/pull/30986))
- Enhanced memory estimation during loading by adding a load memory factor ([30999](https://github.com/milvus-io/milvus/pull/30999))
- Reduced memory consumption for DataNode when dealing with multiple collections ([30991](https://github.com/milvus-io/milvus/pull/30991))
- Optimized the connection manager in the proxy to prevent out-of-memory errors ([31009](https://github.com/milvus-io/milvus/pull/31009))

Bug Fixes

- Fixed search/query failed caused by passing an incorrect context and missing error handling ([30818](https://github.com/milvus-io/milvus/pull/30818))
- Fixed incorrect data types in the output when using the RESTful query interface ([30738](https://github.com/milvus-io/milvus/pull/30738))
- Fixed incorrect disk estimation when loading disk index (or index with mmap enabled) ([30948](https://github.com/milvus-io/milvus/pull/30948))
- Fixed the issue where redundant segments with older versions may never be released ([30953](https://github.com/milvus-io/milvus/pull/30953))
- Fixed the issue where incorrect disk limits may be read ([30966](https://github.com/milvus-io/milvus/pull/30966))
- Fixed a series of issues with the RESTful interface, such as incorrect int64 precision and failed insertion ([30873](https://github.com/milvus-io/milvus/pull/30873))
- Fixed panic during rolling upgrade caused by compatibility issues ([30656](https://github.com/milvus-io/milvus/pull/30656))
- Fixed an issue where flush and compaction were stalled due to uncleared compaction tasks in the Datanode ([30972](https://github.com/milvus-io/milvus/pull/30972))
- Fixed frequent lock contention in the GetCompactionTo operation, which caused channel checkpoints lag ([30965](https://github.com/milvus-io/milvus/pull/30965))
- Fixed the issue where Datanode failed to update channel checkpoints with multiple collections ([31024](https://github.com/milvus-io/milvus/pull/31024) [#31082](https://github.com/milvus-io/milvus/pull/31082))
- Fixed insertion failures caused by frequent contention for locks ([31026](https://github.com/milvus-io/milvus/pull/31026))
- Fixed proxy panic caused by missing error handling ([31086](https://github.com/milvus-io/milvus/pull/31086))