Commit Graph

10037 Commits (817b616eb4c22d433ff46ffeb65a8f40761df073)

Author SHA1 Message Date
cai.zhang 817b616eb4
fix: [2.5]Restore the compacting state for stats task during recovery (#39460)
issue: #39333 

master pr: #39459

---------

Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
2025-01-21 01:03:05 +08:00
aoiasd 15d60c60c2
enhance: [2.5] remain token after authorize (#39455)
relate: https://github.com/milvus-io/milvus/issues/39285

Signed-off-by: aoiasd <zhicheng.yue@zilliz.com>
2025-01-20 20:27:04 +08:00
smellthemoon e4c6ef9d57
enhance: [2.5] serializes the resp body and attach some details (#39401)
issue: #37611 
pr: #39140

Signed-off-by: lixinguo <xinguo.li@zilliz.com>
Co-authored-by: lixinguo <xinguo.li@zilliz.com>
2025-01-20 18:17:10 +08:00
smellthemoon ff53d4f742
enhance: [2.5]add some apis in Restful (#39425)
pr: #39105 
issue: #38709

---------

Signed-off-by: lixinguo <xinguo.li@zilliz.com>
Co-authored-by: lixinguo <xinguo.li@zilliz.com>
2025-01-20 18:15:22 +08:00
congqixia 8e2e828a78
enhance: [2.5] Skip loading pk index for sorted segment in loader (#39438)
Cherry pick from master
pr: #39437

Related to #39339
Previous PR #39389 only skips append index into segment

Also related to #39428

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2025-01-20 18:07:04 +08:00
zhenshan.cao 964000f645
fix: deleted the sealed segment data accidentally (#39422)
issue:https://github.com/milvus-io/milvus/issues/39333
pr: https://github.com/milvus-io/milvus/pull/39421

Signed-off-by: zhenshan.cao <zhenshan.cao@zilliz.com>
2025-01-20 17:49:03 +08:00
Bingyi Sun 6342237ca8
enhance: add workaround for concurrent tantivy reload and directory removal (#39441)
related pr: https://github.com/milvus-io/milvus/pull/39440
issue: https://github.com/milvus-io/milvus/issues/39428

---------

Signed-off-by: sunby <sunbingyi1992@gmail.com>
2025-01-20 17:33:05 +08:00
Bingyi Sun 6f9f840080
fix: Fix panic caused by removing directory (#38622) (#39434)
related pr: https://github.com/milvus-io/milvus/pull/38622
issue: https://github.com/milvus-io/milvus/issues/39428

Signed-off-by: sunby <sunbingyi1992@gmail.com>
2025-01-20 16:13:03 +08:00
sthuang 8c35064d01
enhance: [2.5] clean up logs of empty op key (#39394)
cherry-pick from master: https://github.com/milvus-io/milvus/pull/39335
related: https://github.com/milvus-io/milvus/issues/39334

Signed-off-by: shaoting-huang <shaoting.huang@zilliz.com>
2025-01-20 12:15:10 +08:00
yihao.dai 4464966462
enhance: [2.5] Remove frequent observe log (#39414)
/kind improvement

pr: https://github.com/milvus-io/milvus/pull/39413

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2025-01-20 11:01:10 +08:00
Gao 6d5f174142
enhance: [2.5] add recalls response for restful api (#39409)
issue: #37899 
pr: #39408

Signed-off-by: chasingegg <chao.gao@zilliz.com>
2025-01-20 10:59:10 +08:00
Gao efe76905c5
fix: correct remote centroids path in clustering compaction (#39399)
issue: https://github.com/milvus-io/milvus/issues/39353
pr: #39398

Signed-off-by: chasingegg <chao.gao@zilliz.com>
2025-01-20 10:57:21 +08:00
XuanYang-cn c9b0859b16
fix: [cp25]Record active collections for l0Policy (#39217) (#39383)
By recording the active collection lists, The l0 compaction trigger of
view change and idle won't influence each other.

Also this pr replaces the L0View cache with real L0 segments' change.
Save some memory and make L0 compaction triggers more accurate.

See also: #39187
pr: #39217

Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2025-01-20 10:47:03 +08:00
sparknack 57bcb1bfc0
fix: sparse: add inverted_index_algo and dim_max_score_ratio config (#39362)
issue: #39332
pr: #39358

Signed-off-by: Shawn Wang <shawn.wang@zilliz.com>
2025-01-18 12:17:03 +08:00
congqixia 0c00f94965
fix: [2.5] Skip load extra indexes for sorted segment pk field (#39390)
Cherry-pick from master
pr: #39389
Related to #39339

Extra indexes can be ignored for most cases since sorted pk column
already provided indexing features

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2025-01-17 14:53:03 +08:00
yihao.dai 89a183c7c2
enhance: [2.5] enable task delta cache (#39349)
When there are many segment tasks in the querycoord scheduler, the
traversal in GetSegmentTaskDelta checks becomes time-consuming. This PR
adds caching for segment deltas.

issue: https://github.com/milvus-io/milvus/issues/37630

pr: https://github.com/milvus-io/milvus/pull/39307

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
Co-authored-by: Wei Liu <wei.liu@zilliz.com>
2025-01-17 12:01:03 +08:00
Zhen Ye 5330fc981c
fix: panic when streaming release if using msgstream (#39375)
issue: #39367
pr: #39374

Signed-off-by: chyezh <chyezh@outlook.com>
2025-01-17 11:55:03 +08:00
Bingyi Sun 56cb1683eb
fix: Fix performance issue and use after free bug (#39343)
cherry pick some fixes in https://github.com/milvus-io/milvus/pull/39249

Signed-off-by: sunby <sunbingyi1992@gmail.com>
2025-01-17 11:51:03 +08:00
Spade A 47e410feb1
fix: [2.5] fix assignment operator in AssertInfo to comparison operator (#39351)
cp https://github.com/milvus-io/milvus/pull/39347
issue: #39346

Signed-off-by: SpadeA-Tang <tangchenjie1210@gmail.com>
2025-01-17 11:27:02 +08:00
Zhen Ye 0391b8c7d3
fix: update rg failed when alter database (#39356)
issue: #36933
pr: #39355

Signed-off-by: chyezh <chyezh@outlook.com>
2025-01-17 10:37:03 +08:00
yihao.dai b69994272f
enhance: [2.5] Limit the maximum number of segments restored and fail the job if saving the binlog fails (#39359)
1. Limit the maximum number of restored segments to 1024.
2. Fail the import job if saving binlog fails.
3. Fail the import job if saving the import task fails to prevent
repeatedly generating dirty importing segments.

issue: https://github.com/milvus-io/milvus/issues/39331

pr: https://github.com/milvus-io/milvus/pull/39344

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2025-01-17 10:27:04 +08:00
yihao.dai 6773fb10a8
enhance: [2.5] Read metadata concurrently to accelerate recovery (#38900)
Read metadata such as segments, binlogs, and partitions concurrently at
the collection level.

issue: https://github.com/milvus-io/milvus/issues/37630

pr: https://github.com/milvus-io/milvus/pull/38403

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2025-01-16 17:53:01 +08:00
yihao.dai 9d2a0e775c
fix: [2.5] Fix slow dist handle and slow observe (#38905)
1. Provide partition&channel level indexing in the collection target.
2. Make SegmentAction not wait for distribution.
3. Remove scheduler and target manager mutex
4. Optimize logging to reduce CPU overhead.

issue: https://github.com/milvus-io/milvus/issues/37630

pr: https://github.com/milvus-io/milvus/pull/38566

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2025-01-16 17:07:02 +08:00
Xianhui Lin 448eb3b63a
fix: [2.5]bulkimport binlogs skip diskquota check (#39319)
bulkimport binlogs skip diskquota check
issue:https://github.com/milvus-io/milvus/issues/38845
pr: https://github.com/milvus-io/milvus/pull/39315

Signed-off-by: Xianhui.Lin <xianhui.lin@zilliz.com>
2025-01-16 15:33:05 +08:00
yihao.dai 29dad64341
fix: [2.5] Fix consume blocked due to too many consumers (#38915)
This PR limits the maximum number of consumers per pchannel to 10 for
each QueryNode and DataNode.

issue: https://github.com/milvus-io/milvus/issues/37630

pr: https://github.com/milvus-io/milvus/pull/38455

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2025-01-16 15:19:03 +08:00
SimFG 4cb47b3f67
feat: [2.5] root privileges can be customized (#39324)
- issue: #39184
- pr: #39191

Signed-off-by: SimFG <bang.fu@zilliz.com>
2025-01-16 14:27:11 +08:00
yihao.dai c945efa853
enhance: [2.5] Remove mutex from SegmentManger in querynode (#39051) (#39282)
Remove mutex from SegmentManger in querynode to prevent mutex
contention.

issue: https://github.com/milvus-io/milvus/issues/37630

pr: https://github.com/milvus-io/milvus/pull/39051

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2025-01-16 13:27:02 +08:00
yihao.dai c741b8be2b
fix: [2.5] Remove frequently updating metric to avoid mutex contention (#38778)
issue: https://github.com/milvus-io/milvus/issues/37630

Reduce the frequency of `updateIndexTasksMetrics` to avoid holding the
mutex for long periods.

pr: https://github.com/milvus-io/milvus/pull/38775

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2025-01-16 11:51:02 +08:00
smellthemoon 1602390734
enhance: support null expr(#38772) (#38931)
pr: #38772 
issue: #31728

---------

Signed-off-by: lixinguo <xinguo.li@zilliz.com>
Co-authored-by: lixinguo <xinguo.li@zilliz.com>
2025-01-16 10:55:07 +08:00
wei liu 76ed552b00
enhance: Add logs for check health failed (#39208) (#39302)
pr: #39208

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2025-01-16 10:31:04 +08:00
congqixia 21df11b37c
enhance: [2.5] Add try-catch and return CStatus for NewCollection (#39279) (#39303)
Cherry pick from master
pr: #39279 
Related to #28795

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2025-01-16 10:13:06 +08:00
Spade A 1d9788e4bc
fix: [2.5] enable to build index with single segment (#39272)
issue: #39232
cherry-pick https://github.com/milvus-io/milvus/pull/39233

---------

Signed-off-by: SpadeA-Tang <tangchenjie1210@gmail.com>
2025-01-16 09:59:02 +08:00
congqixia b658467107
enhance: [2.5] Unify LoadStateLock RLock & PinIf (#39206) (#39255)
Cherry-pick from master
pr: #39206 #39308
Related to #39205

This PR merge `RLock` & `PinIfNotReleased` into `PinIf` function
preventing segment being released before any Read operation finished.

---------

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2025-01-16 00:33:03 +08:00
yihao.dai 9b916f2cf8
enhance: [2.5] Skip creating partition rate limiters when not enable (#38911)
issue: https://github.com/milvus-io/milvus/issues/37630

pr: https://github.com/milvus-io/milvus/pull/38816

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2025-01-15 22:19:02 +08:00
Gao dd44a58381
enhance: [2.5] pass partition key scalar info if enable for vector mem index (#39245)
issue: #34332 
pr: #39123

---------

Signed-off-by: chasingegg <chao.gao@zilliz.com>
2025-01-15 21:45:01 +08:00
congqixia 2fe245f918
fix: [2.5] Add index param duplication check (#39289) (#39304)
Cherry-pick from master
pr: #39289
Related to #39288

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2025-01-15 21:04:06 +08:00
yihao.dai 8bf0f0b921
fix: [2.5] Fix rootcoord meta mutex contention (#38902)
RootCoord meta uses copy-on-write, allowing the removal of unnecessary
copies.

issue: https://github.com/milvus-io/milvus/issues/37630

pr: https://github.com/milvus-io/milvus/pull/38799

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2025-01-15 20:11:08 +08:00
smellthemoon 2439278f93
enhance: add some apis in Restful(#38733) (#38875)
issue: #38709 
pr: #38733  #38979

---------

Signed-off-by: lixinguo <xinguo.li@zilliz.com>
Co-authored-by: lixinguo <xinguo.li@zilliz.com>
2025-01-15 15:19:00 +08:00
yihao.dai 5b0bb4c04e
enhance: [2.5] Reduce memory usage of BF in DataNode and QueryNode (#38913)
1. DataNode: Skip generating BF during the insert phase (BF will be
regenerated during the sync phase).
2. QueryNode: Skip generating or maintaining BF for growing segments;
deletion checks will be handled in the segcore.

issue: https://github.com/milvus-io/milvus/issues/37630

pr: https://github.com/milvus-io/milvus/pull/38129

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2025-01-15 13:59:00 +08:00
Zhen Ye 74246d8167
fix: wrong streaming mockery package name (#39261)
issue: #39095
pr: #39260

Signed-off-by: chyezh <chyezh@outlook.com>
2025-01-15 11:44:59 +08:00
Chun Han 4c91e05a5d
enhance: fix inconsistenty of alias and db for query iterator(#39045) (#39248)
related: #39045
pr: https://github.com/milvus-io/milvus/pull/39216

Signed-off-by: MrPresent-Han <chun.han@gmail.com>
Co-authored-by: MrPresent-Han <chun.han@gmail.com>
2025-01-15 10:36:59 +08:00
cai.zhang 6816ee4cf5
fix: [2.5] Record a map to avoid repeatedly traversing the CompactionFrom (#38926)
issue: #38811 

master pr: #38925

Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
2025-01-15 10:31:00 +08:00
yihao.dai 2e4a1052aa
enhance: [2.5] Reduce mutex contention in datacoord meta (#38904)
1. Using secondary index to avoid retrieving all segments at
GetSegmentsChanPart.
2. Perform batch SetAllocations to reduce the number of times the meta
lock is acquired.

issue: https://github.com/milvus-io/milvus/issues/37630

pr: https://github.com/milvus-io/milvus/pull/38219

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2025-01-15 00:55:00 +08:00
cai.zhang 4270174899
fix: [2.5] Add scalar index engine version for compatibility (#39236)
issue: #39203 

master pr: #39204

---------

Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
2025-01-14 21:01:01 +08:00
yihao.dai b91c0a8079
enhance: [2.5] Optimize GetLocalDiskSize and segment loader mutex (#38907)
1. Make the segment loader lock protect only the resource.
2. Optimize GetDiskUsage to avoid excessive overhead.

issue: https://github.com/milvus-io/milvus/issues/37630

pr: https://github.com/milvus-io/milvus/pull/38599

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2025-01-14 20:43:06 +08:00
wei liu 51994158d9
fix: channel unbalance during stopping balance progress (#38971) (#39200)
issue: #38970
pr: #38971
cause the stopping balance channel still use the row_count_based policy,
which may causes channel unbalance in multi-collection case.

This PR impl a score based stopping balance channel policy.

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2025-01-14 18:25:00 +08:00
wei liu 4fd56e4773
fix: Prevent leader checker from generating excessive duplicate leader tasks (#39000) (#39160)
issue: #39001
pr: #39000
Background:
Segment Load Version: Each segment load request assigns a timestamp as
its version. When multiple copies of a segment are loaded on different
QueryNodes, the leader checker uses this version to identify the latest
copy and updates the routing table in the leader view to point to it.
Delegator Router Version: When a delegator builds a route to a QueryNode
that has loaded a segment, it also records the segment's version.

Router Table Update Logic: If the leader checker detects that the
version of a segment in the routing table does not match the version in
the worker, it updates the routing table to point to the QueryNode with
the latest version. Additionally, it updates the segment's load version
in the QueryNode during this process.

Issue:
When a channel is undergoing load balancing, the leader checker may sync
the routing table to a new delegator. This sync operation modifies the
segment's load version, which invalidates the routing in the old
delegator. Subsequently, the leader checker updates the routing table in
the old delegator, breaking the routing in the new delegator. This cycle
continues, causing repeated updates and inconsistencies.

Fix:
This PR introduces two changes to address the issue:
1. Use NodeID to verify whether the delegator's routing table needs an
update, avoiding unnecessary modifications.
2. Ensure compatibility by using the latest segment's load version as
the version recorded in the routing table.

These changes resolve the cyclic updates and prevent the leader checker
from generating excessive duplicate tasks, ensuring routing stability
across delegators during load balancing.

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2025-01-14 18:11:06 +08:00
congqixia b4ddf746c1
enhance: [2.5][bitset] extend op_find() to be able to search both 0 and 1 (#39237)
Cherry-pick from master
pr: #39176 #39229
issue: #39124

`bitset::find_first()` and `bitset::find_next()` now accept one more
parameter, which allows to search for `0` bit instead of `1` bit

---------

Signed-off-by: Alexandr Guzhva <alexanderguzhva@gmail.com>
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
Co-authored-by: Alexander Guzhva <alexanderguzhva@gmail.com>
2025-01-14 17:59:00 +08:00
aoiasd e6ac2fe063
enhance: [Cherry-pick] report auth log to hook when grpc authenticate failed (#39192)
relate: https://github.com/milvus-io/milvus/pull/38943

Signed-off-by: aoiasd <zhicheng.yue@zilliz.com>
2025-01-14 17:47:04 +08:00
sthuang 4a47f760b3
fix: [2.5] rbac custom group privilege level check (#39164) (#39224)
cherry-pick from master: https://github.com/milvus-io/milvus/pull/39164
related: https://github.com/milvus-io/milvus/issues/39086

Signed-off-by: shaoting-huang <shaoting.huang@zilliz.com>
2025-01-14 16:50:59 +08:00