145 Commits

Author SHA1 Message Date
yah01
57397b1307
enhance: add new LRU cache impl (#30360)
- remove  the unused LRU cache
- add new LRU cache impl which wraps github.com/karlseguin/ccache

related #30361

---------

Signed-off-by: yah01 <yang.cen@zilliz.com>
2024-02-27 20:58:40 +08:00
zhagnlu
e8a6f1ea2b
fix: erase pk empty check when pk index replace raw data (#30432)
#30350

Signed-off-by: luzhang <luzhang@zilliz.com>
Co-authored-by: luzhang <luzhang@zilliz.com>
2024-02-07 14:56:47 +08:00
cqy123456
5449e862d5
fix: safety access unordered_map and remove some useless code excute (#30504)
issue: https://github.com/milvus-io/milvus/issues/30358 and
https://github.com/milvus-io/milvus/issues/30491

Signed-off-by: cqy123456 <qianya.cheng@zilliz.com>
2024-02-05 22:03:09 +08:00
cqy123456
74cfba0249
enhance:limit binlog index rows num (#30173)
issue: https://github.com/milvus-io/milvus/issues/27678
also relate issue: https://github.com/milvus-io/milvus/issues/30065

Signed-off-by: cqy123456 <qianya.cheng@zilliz.com>
2024-01-29 19:49:02 +08:00
xige-16
e9fdd2475d
fix: fix searchPlan metricType modified concurrently (#30227)
issue: #30225
/kind bug
Signed-off-by: xige-16 <xi.ge@zilliz.com>

---------

Signed-off-by: xige-16 <xi.ge@zilliz.com>
2024-01-26 14:03:09 +08:00
yihao.dai
c02fb64ad6
enhance: Allows proactive warming up of chunk cache (#30182)
Allows proactive warming up of chunk cache. Original vector data will be
asynchronously loaded into the chunk cache during the load process. It
has the potential to significantly reduce query/search latency for a
certain duration after the load, albeit with a concurrent increase in
disk usage.

issue: https://github.com/milvus-io/milvus/issues/30181

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-01-25 19:55:39 +08:00
yah01
f542bdbf3c
enhance: calc the accurate mem size of segment (#30093)
this stats the real memory size of segment, also reduces the memory
usage in mmap mode
resolve #30095

Signed-off-by: yah01 <yang.cen@zilliz.com>
2024-01-19 12:32:53 +08:00
xige-16
fa7cf587b0
enhance: Opt metric type does not match error message (#29927)
issue: #29791 
/kind improvement
Signed-off-by: xige-16 <xi.ge@zilliz.com>

Signed-off-by: xige-16 <xi.ge@zilliz.com>
2024-01-17 20:25:03 +08:00
yah01
1185e4dcd5
fix: written file size is over the int32 range and raises error (#30057)
we sum the total data size in int32, which could lead to an overflow
error
related #30056

Signed-off-by: yah01 <yang.cen@zilliz.com>
2024-01-17 16:42:54 +08:00
yah01
6c477ce3a7
enhance: optimize the loading strategy (#29910)
as we have the pool size limit so we don't need to limit the concurrency
manually

Signed-off-by: yah01 <yang.cen@zilliz.com>
2024-01-12 14:26:50 +08:00
yah01
aba2656e68
fix: missing field data after appending scalar index to loaded segment (#29912)
related #29843

Signed-off-by: yah01 <yang.cen@zilliz.com>
2024-01-12 14:04:54 +08:00
Xu Tong
e429965f32
Add float16 approve for multi-type part (#28427)
issue:https://github.com/milvus-io/milvus/issues/22837

Add bfloat16 vector, add the index part of float16 vector.

Signed-off-by: Writer-X <1256866856@qq.com>
2024-01-11 15:48:51 +08:00
xige-16
9702cef2b5
feat: Support multiple vector search (#29433)
issue #25639 

Signed-off-by: xige-16 <xi.ge@zilliz.com>

Signed-off-by: xige-16 <xi.ge@zilliz.com>
2024-01-08 15:34:48 +08:00
cai.zhang
5dc300c4a9
fix: Fix bug for pk index doesn't have raw data (#29711)
issue: #29697

Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
2024-01-07 19:36:48 +08:00
MrPresent-Han
9e2e7157e9
feat: support search_group_by for milvus(#25324) (#28983)
related: #25324

Search GroupBy function, used to aggregate result entities based on a
specific scalar column.
several points to mention:

1. Temporarliy, the whole groupby is implemented separated from
iterative expr framework **for the first period**
2. In the long term, the groupBy operation will be incorporated into the
iterative expr framework:https://github.com/milvus-io/milvus/pull/28166
3. This pr includes some unrelated mocked interface regarding alterIndex
due to some unworth-to-mention reasons. All these un-associated content
will be removed before the final pr is merged. This version of pr is
only for review
4. All other related details were commented in the files comparison

Signed-off-by: MrPresent-Han <chun.han@zilliz.com>
2024-01-05 15:50:47 +08:00
yah01
aef483806d
enhance: improve the segcore logs (#29372)
- remove the streaming logging
- refine existing logs

fix #29366

---------

Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-12-23 21:52:43 +08:00
chyezh
be87c18b44
fix: fixup data race at generate binlog index (#29370)
issue: #29339

Signed-off-by: chyezh <ye.zhen@zilliz.com>
2023-12-21 14:58:49 +08:00
zhagnlu
a602171d06
enhance: Refactor runtime and expr framework (#28166)
#28165

Signed-off-by: luzhang <luzhang@zilliz.com>
Co-authored-by: luzhang <luzhang@zilliz.com>
2023-12-18 12:04:42 +08:00
cai.zhang
fb089cda8b
enhance: Load raw data while scalar index doesn't have raw data (#28888)
issue: #28886

Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
2023-12-06 20:36:36 +08:00
Bingyi Sun
36f69ea031
feat: integrate storagev2 in building index of segcore (#28768)
issue: https://github.com/milvus-io/milvus/issues/28655

---------

Signed-off-by: sunby <sunbingyi1992@gmail.com>
2023-12-05 16:48:54 +08:00
yihao.dai
f5856812a2
fix: Fix get binary vector from chunk cache (#28866)
The way of getting binary vector size is wrong. This PR will fix it.

issue: https://github.com/milvus-io/milvus/issues/28865

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2023-12-01 14:40:32 +08:00
cqy123456
3b1b14dd78
fix: update binlog index memory uasge before loading segments (#28528)
issue: #27678 
when interimIndex = true, memory predict should be update with the
memory usage of binlog index build process.

Signed-off-by: cqy123456 <qianya.cheng@zilliz.com>
2023-11-29 16:42:27 +08:00
yah01
f7d2ab6677
enhance: reduce 1x copy for variable length field while retrieving (#28345)
- Reduce 1x copy for varchar/string/JSON/array types while retrieving
- Reduce 1x copy for int8/int16 while retrieving

Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-11-15 18:08:20 +08:00
MrPresent-Han
836f300536
support skip-index based on chunk-metrics to accelerate expr filter(#27925) (#28297)
related: #27925

Signed-off-by: MrPresent-Han <chun.han@zilliz.com>
2023-11-15 11:20:19 +08:00
Xu Tong
8ec85f5f4c
Add template for VectorMemIndex (#28324)
Signed-off-by: Writer-X <1256866856@qq.com>
2023-11-11 13:20:22 +08:00
yah01
30847cad3e
Handle exception while loading (#28304)
Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-11-09 17:59:12 +08:00
yah01
863e26969a
Reduce 1x memory copy for retrieving data (#28106)
Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-11-06 17:48:19 +08:00
yah01
dc89730a50
Support collection-level mmap control (#26901)
Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-11-02 23:52:16 +08:00
cqy123456
4fbe3c9142
replace loaded binlog with binlog index for search performance (#27673)
Signed-off-by: cqy123456 <qianya.cheng@zilliz.com>
2023-11-01 02:20:15 +08:00
yihao.dai
ab6b0103a3
Get vector concurrently (#27838)
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2023-10-30 15:44:12 +08:00
yah01
f212158d61
Fix delete records timestamp may be reordered (#27941)
Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-10-27 10:24:10 +08:00
Enwei Jiao
b80a3e19d3
Add code for PanicInfo (#27364)
Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>
2023-09-27 12:01:28 +08:00
yah01
93e2eb78c9
Delete only if primary keys exist (#25292)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-09-20 19:03:25 +08:00
cai.zhang
a362bb1457
Support array datatype (#26369)
Signed-off-by: cai.zhang <cai.zhang@zilliz.com>
2023-09-19 14:23:23 +08:00
yihao.dai
bb6711f28c
Add ChunkCache: support get vector from storage (#26142)
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2023-09-15 10:21:20 +08:00
Enwei Jiao
0afdfdb9af
Remove other Exceptions, keeps SegcoreError only (#27017)
Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>
2023-09-14 14:05:20 +08:00
yah01
3203ce1654
Reduce copy while retrieving primary keys (#26616)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-09-11 21:31:18 +08:00
Enwei Jiao
c3f15c6b95
Refactor duplicate error class into one place (#26985)
Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>
2023-09-11 20:43:17 +08:00
Xu Tong
9166011c4a
Add float16 vector (#25852)
Signed-off-by: Writer-X <1256866856@qq.com>
2023-09-08 10:03:16 +08:00
MrPresent-Han
8330c18dc9
add log for loading segment(#26564) (#26640)
/kind improvement

Signed-off-by: MrPresent-Han <chun.han@zilliz.com>
2023-08-31 12:03:00 +08:00
MrPresent-Han
7d5a4b2994
add more event for segcore search(#26277) (#26688)
Signed-off-by: MrPresent-Han <chun.han@zilliz.com>
2023-08-30 14:15:01 +08:00
yah01
b475f25042
Remove invalid offset check while filling data (#26666)
Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-08-30 09:52:27 +08:00
yah01
9605c03c3c
Fix the number of rows of column not correct (#26347)
Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-08-16 13:35:33 +08:00
yah01
127c23d999
Check data consistency after loading (#26312)
Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-08-14 09:01:32 +08:00
xige-16
1055c90456
Add default retrieve limit (#24782)
Signed-off-by: xige-16 <xi.ge@zilliz.com>
2023-08-10 14:11:15 +08:00
yah01
300fef446b
Enable mmap for vector index (#25877)
Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-08-10 13:59:15 +08:00
smellthemoon
45b8e1b1d9
Fix log (#26165)
Signed-off-by: lixinguo <xinguo.li@zilliz.com>
Co-authored-by: lixinguo <xinguo.li@zilliz.com>
2023-08-08 15:35:07 +08:00
MrPresent-Han
5634ba777d
add new threadpool with various priority to avoid deadlock(#25781) (#26028)
Signed-off-by: MrPresent-Han <chun.han@zilliz.com>
2023-08-03 09:31:07 +08:00
Jiquan Long
5c1f79dc54
Push down the limit operator to segcore (#25959)
Signed-off-by: longjiquan <jiquan.long@zilliz.com>
2023-08-01 20:29:05 +08:00
yah01
9618bd9b42
Set channel capacity before consuming it (#25895)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-07-26 17:35:01 +08:00