567 Commits

Author SHA1 Message Date
Zhen Ye
5d2f454ce4
enhance: add multiply factor when loading index (#38721)
issue: #38715
pr: #38716

Signed-off-by: chyezh <chyezh@outlook.com>
2024-12-25 10:50:58 +08:00
congqixia
6f573f64a5
enhance: [2.4] Remove levelZeroMut totally (#38473) (#38495)
Cherry pick from master
pr: #38473
The level zero mutex could be remove since all operations are guarded by
segment manager mutex

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-12-17 10:18:45 +08:00
SimFG
df73f93126
enhance: [2.4] pick some master improvements to 2.4 branch (#38128)
- issue: #38127

master pr list:
- #37759
- #37835
- #37845
- #37874
- #37894
- #37969
- #37983
- #38005
- #38035

---------

Signed-off-by: SimFG <bang.fu@zilliz.com>
2024-12-13 10:24:45 +08:00
Gao
d7e926f1cb
enhance: [2.4] support recall estimation (#38064)
issue: #37899 
pr: #38017

---------

Signed-off-by: chasingegg <chao.gao@zilliz.com>
2024-12-11 20:40:48 +08:00
wei liu
0c442b6fd6
fix: Datanode stop progress stuck at writer buffer memory check (#38274) (#38287)
issue: #38273
pr: #38274 #38278

---------

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2024-12-09 17:36:42 +08:00
congqixia
f888267d37
enhance: [2.4] Fill version for load delta request (#38212) (#38228)
Cherry-pick from master
pr: #38212
Version is needed for load delta request in case of false alarm warning
about version go backward

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-12-05 14:08:41 +08:00
jaime
319f5494cd
enhance: optimize CPU usage for CheckHealth requests (#35595)
issue: #35563
pr: #35589

Signed-off-by: jaime <yun.zhang@zilliz.com>
2024-12-04 14:26:41 +08:00
wei liu
b29237e5d5
enhance: Add collection id to search request count metrics (#38069) (#38144)
pr: #38069 #38167

---------

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2024-12-03 18:16:39 +08:00
congqixia
c326a52370
fix: [2.4] Use correct policy merging growing&l0 and add unit tests (#37950) (#37968)
Cherry-pick from master
pr: #37950
Related to #37574

---------

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-11-28 14:22:37 +08:00
Gao
165afbba91
enhance: support retry search when topk is reduced and result not enough (#37093)
issue: #35576 
pr: #35645

---------

Signed-off-by: chasingegg <chao.gao@zilliz.com>
2024-11-28 10:12:37 +08:00
wei liu
e63a2f3559
fix: unstable integration test caused by paramtable.GetNodeID (#37910)
issue: #37908
pr: #37909
cause paramtable is global single instance, which cause
paramtable.GetNodeID may return wrong server id in integration test.

This PR use node.GetNodeID to replace paramtable.GetNodeID

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2024-11-22 10:14:33 +08:00
congqixia
876e06b862
fix: [2.4] Load l0 delta for growings when using RemoteLoad (#37772)
Cherry-pick from master
pr: #37771
Related to #37574

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-11-18 20:26:31 +08:00
wei liu
2a4f54cd4f
fix: L0 segment has been loaded to worker during channel balance (#37758)
issue: https://github.com/milvus-io/milvus/issues/37703
pr: https://github.com/milvus-io/milvus/pull/37748

---------

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2024-11-18 17:00:32 +08:00
XuanYang-cn
d23da2db4f
fix: [cp24]Correct varchar primarykey size calculation (#37619)
See also: #37582
pr: #37617

---------

Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2024-11-14 14:16:38 +08:00
congqixia
d073f322a4
enhance: [2.4] Add cgo call metrics for load/write API (#37405) (#37627)
Cherry-pick from master
pr: #37405

Cgo API cost is not observerable since not metrics is related to them.
This PR add metrics for some sync cgo call related to load & write

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-11-13 13:58:30 +08:00
cai.zhang
3456e241ac
fix: [2.4]Fix the bug that retrieved from wrong field for L0 segments (#37599)
issue: #37574 

master pr: #37598

Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
2024-11-12 19:02:31 +08:00
congqixia
ce7d4090f1
enhance: [2.4] Move forward l0 logic out of delta lock (#37340)
Cherry pick from master
pr: #37337
Related to #35303

`deleteMut` shall be protecting streaming delete buffer, forward l0
could be move out of the rlock section to reduce tsafe impact from
loading segments.

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-11-01 14:42:22 +08:00
congqixia
a2a51c489e
fix: [2.4] Check resource when loading deltalogs (#37195) (#37263)
Cherry pick from master
pr: #37195
Related to #36887

`LoadDeltaLogs` API did not check memory usage. When system is under
high delete load pressure, this could result into OOM quit.

This PR add resource check for `LoadDeltaLogs` actions and separate
internal deltalog loading function with public one.

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-10-30 11:54:41 +08:00
aoiasd
8370caa4a6
enhance: [Cherry-pick]Add collection name label for some metric (#36951) (#37159)
pr: https://github.com/milvus-io/milvus/pull/36951

Signed-off-by: aoiasd <zhicheng.yue@zilliz.com>
2024-10-29 17:38:22 +08:00
congqixia
3d1e81fb31
fix: [2.4] Use singleton delete pool and avoid goroutine leakage (#37225)
Cherry-pick from master
pr: #37220
Related to #36887

Previously using newly create pool per request shall cause goroutine
leakage. This PR change this behavior by using singleton delete pool.
This change could also provide better concurrency control over delete
memory usage.

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-10-29 14:44:23 +08:00
cai.zhang
9c0f59488a
feat: [cherry-pick]The expression supports filling elements through templates (#37058)
issue: #36672 

master pr: #37033 

milvus-proto pr: https://github.com/milvus-io/milvus-proto/pull/332

Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
2024-10-28 15:17:30 +08:00
congqixia
223badc482
fix: [2.4] Ref collection meta when load l0 segment meta only (#37179)
Cherry pick from master
pr: #37178
Related to #37177

Previous PR #37160

Collection meta is not ref-ed when loading l0 segment in `RemoteLoad`
policy, which cause collection meta release when lots of l0 segment
released.

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-10-28 14:07:30 +08:00
congqixia
9d37ade24f
enhance: [2.4] Make skip load work for all branches (#37161)
Cherry-pick from master
pr: #37160
Related to #37112

Skip load logic used to work only when there is multiple segment load
info entires in load request. In continous delete case, delegator still
loads l0 segment, which occupies lot of memory.

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-10-25 22:11:30 +08:00
yihao.dai
da897e41f4
fix: Fix collection leak in querynode (#37061) (#37079)
Unref the removed L0 segment count.

issue: https://github.com/milvus-io/milvus/issues/36918

pr: https://github.com/milvus-io/milvus/pull/37061

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-10-25 18:19:39 +08:00
congqixia
6bc8aba17f
enhance: [2.4] Batch forward delete when using DirectForward (#37076) (#37107)
Cherry pick from master
pr: #37076
Related #36887

DirectFoward streaming delete will cause memory usage explode if the
segments number was large. This PR add batching delete API and using it
for direct forward implementation.

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-10-25 11:53:29 +08:00
congqixia
79891f047d
enhance: [2.4]Skip load delta data in delegater when using RemoteLoad (#37082) (#37112)
Cherry-pick from master
pr: #37082
Related to #35303

Delta data is not needed when using `RemoteLoad` l0 forward policy. By
skipping load delta data, memory pressure could be eased if l0 segment
size/number is large.

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-10-25 11:41:30 +08:00
congqixia
3db137f4ad
enhance: [2.4] Add metrics for querynode delete buffer info (#37081) (#37097)
Cherry pick from master
pr: #37081
Related to #35303

This PR add metrics for querynode delegator delete buffer information,
which is related to dml quota logic.

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-10-24 16:21:37 +08:00
congqixia
7eba3aa67e
fix: [2.4] Pass full field list when partial load enabled (#37053) (#37063)
Cherry-pick from master
pr: #37053

Related to #37038

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-10-23 11:03:28 +08:00
congqixia
7acf1d53c1
enhance: [2.4] Preallocate delete data slice to avoid growslice (#37044)
Rewritten based on master pr
pr: #37043

Related to #36887

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-10-22 14:15:28 +08:00
wei liu
1dcc393e54
fix: Query node panic during sending rpc to worker (#36975) (#36988)
issue: #36976
pr: #36975

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2024-10-21 10:15:26 +08:00
yihao.dai
8923936c9a
enhance: Support memory mode chunk cache (#35347) (#35836)
Chunk cache supports loading raw vectors into memory.

issue: https://github.com/milvus-io/milvus/issues/35273

pr: https://github.com/milvus-io/milvus/pull/35347

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-10-18 17:03:25 +08:00
cqy123456
6934e8da3a
enhance: [2.4]use growingMmapEnabled to control the behavior of interim index, not vectorField (#36391)
issue: https://github.com/milvus-io/milvus/issues/36392
related pr: https://github.com/milvus-io/milvus/pull/36500

Signed-off-by: cqy123456 <qianya.cheng@zilliz.com>
2024-10-17 20:23:25 +08:00
congqixia
dfe27ebf35
fix: [2.4] Direct forward delta exclude l0 segments (#36899) (#36914)
Cherry-pick from master
pr: #36899

Related to #36887

Forward delete to L0 segment will return error and mark l0 segment
offline causing delegator unserviceable

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-10-17 14:11:29 +08:00
congqixia
877e9ad450
enhance: [2.4] Fill start pos & level for growing segment (#36888) (#36911)
Cherry-pick from master
pr: #36888

Start position & level info is missing for growing segment loaded in
watch dml channel operation.

Level is important for metrics and start position is crucial for growing
exclude logic.

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-10-17 14:09:32 +08:00
yihao.dai
604e346585
enhance: Enhance segment log (#36848) (#36849)
/kind improvement

pr: https://github.com/milvus-io/milvus/pull/36848

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-10-15 20:43:30 +08:00
SimFG
548f8e80c3
enhance: [2.4] the estimate method when loading the collection (#36728)
- pr: #36307
- issue: #36530

Signed-off-by: SimFG <bang.fu@zilliz.com>
Signed-off-by: xianliang.li <xianliang.li@zilliz.com>
2024-10-11 10:20:45 +08:00
yihao.dai
9cb5396cf6
enhance: Use common gc config (#36668) (#36670)
Use the GC config from `common` and remove the GC config from
`queryNode`.

issue: https://github.com/milvus-io/milvus/issues/36667

pr: https://github.com/milvus-io/milvus/pull/36668

related pr: https://github.com/milvus-io/milvus/pull/34949

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-10-09 19:49:20 +08:00
congqixia
3a80d1f602
enhance: [2.4] Add streaming forward policy switch for delegator (#36330) (#36712)
Cherry pick from master
pr: #36330
Related to #35303

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-10-09 17:41:20 +08:00
congqixia
9073f6281e
fix: [2.4] Add defer Unpin when error happens (#36620) (#36665)
Cherry-pick from master
pr: #36620
Resolves: #36619

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-10-08 18:29:20 +08:00
wei liu
2428adea3b
enhance: Enable balance on querynode with different mem capacity (#36466) (#36625)
issue: #36464
pr: #36466
This PR enable balance on querynode with different mem capacity, for
query node which has more mem capactity will be assigned more records,
and query node with the largest difference between assignedScore and
currentScore will have a higher priority to carry the new segment.

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2024-09-30 18:11:18 +08:00
SimFG
a00523f0fd
fix: metric type error when the collection has two vec field (#36473)
- issue: #36395

Signed-off-by: SimFG <bang.fu@zilliz.com>
2024-09-24 20:03:14 +08:00
congqixia
fa6354f6df
enhance: [skip e2e][2.4] Add unittest for reducing duplicated pk from multi segments (#36433) (#36460)
Cherry-pick from master
pr: #36433
Related to #35505 #36362

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-09-24 18:09:14 +08:00
wei liu
d421effb03
fix: fix search/query/count may access same growing and sealed segment (#36258) (#36288)
issue: #36257
pr: #36258
during syncTargetVersion, sealed segment should be excluded, to avoid
it's growing segment be conusmed from stream again.

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2024-09-18 22:21:12 +08:00
SimFG
95e47bfcf8
fix: force to set the metric type in the search request (#36279)
- issue: #35960
- pr: #35962

Signed-off-by: SimFG <bang.fu@zilliz.com>
2024-09-18 19:21:11 +08:00
congqixia
f7e4db943c
fix: [2.4] overwrite correct selection when pk duplicated (#35826) (#36274)
Cherry-pick from master
pr: #35826
Related to #35505

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-09-14 14:27:07 +08:00
congqixia
13d443eb2e
enhance: [2.4] Add L0 forward policy to support remote load (#36189) (#36208)
Cherry-pick from master
pr: #36189
Related to #35303

This PR add a param item to support change l0 forward behavior from bf
filtering and forward to remote load.

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-09-12 19:09:08 +08:00
XuanYang-cn
64e109d155
fix: [cp]Change deltalog memory estimation factor to one (#36035)
See also: #36031
pr: #36033

Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2024-09-06 18:09:05 +08:00
XuanYang-cn
54ec290109
enhance: [cp]Remove too frequent logs in Delete (#35981)
pr: #35980

Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2024-09-06 10:47:13 +08:00
congqixia
da0bc22a5f
enhance: [2.4] Add delete buffer related quota logic (#35918) (#35997)
Cherry pick from master
pr: #35128 #35918
See also #35303

---------

Signed-off-by: aoiasd <zhicheng.yue@zilliz.com>
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
Co-authored-by: aoiasd <45024769+aoiasd@users.noreply.github.com>
2024-09-05 16:43:06 +08:00
jaime
2c1fa50412
enhance: remove cooling off in rate limiter for read requests (#35936)
issue: #35934
pr: #35935

Signed-off-by: jaime <yun.zhang@zilliz.com>
2024-09-04 14:39:10 +08:00