Commit Graph

982 Commits (f7a59766df42c7e448e841920625fed4453e2e8c)

Author SHA1 Message Date
congqixia fe2f34d76f
fix: Use server ctx instead of loopCtx for datacoord LivenessCheck (#31691)
See also #31689

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-03-29 10:37:11 +08:00
Bingyi Sun 3d66670619
fix: add lazy load field to mark segment load type (#31591)
issue: https://github.com/milvus-io/milvus/issues/31673

---------

Signed-off-by: sunby <sunbingyi1992@gmail.com>
2024-03-28 11:23:10 +08:00
SimFG b1a1cca10b
feat: add more operation detail info for better allocation (#30438)
issue: #30436

---------

Signed-off-by: SimFG <bang.fu@zilliz.com>
2024-03-28 06:33:11 +08:00
yihao.dai 9a13b9822f
enhance: Return more fields in import progress response (#31539)
Return more fields in import progress response, include importedRows and
totalRows. Additionally, ensure compatibility with the old import
progress response by retaining fields of create timestamp and row count.

issue: https://github.com/milvus-io/milvus/issues/31448
https://github.com/milvus-io/milvus/issues/31237
https://github.com/milvus-io/milvus/issues/28521

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-03-24 21:57:06 +08:00
yihao.dai 0fe5e90e8b
enhance: Remove import v1 (#31403)
Remove all code and logic related to import v1.

issue: https://github.com/milvus-io/milvus/issues/28521

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-03-22 15:29:09 +08:00
congqixia 5c5f53d11b
fix: Check nodeID before update channel checkpoint (#31473)
See also #31470

This PR adds nodeID assignment verification before updating channel
checkpoints.

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-03-22 10:07:06 +08:00
Patrick Weizhi Xu 982dd2834b
enhance: add materialized view search info (#30888)
issue: #29892 

This PR
1. Pass Materialized View (MV) search information obtained from the
expression parsing planning procedure to Knowhere. It only performs when
MV is enabled and the partition key is involved in the expression. The
search information includes:
1. Touched field_id and the count of related categories in the
expression. E.g., `color == red && color == blue` yields `field_id ->
2`.
2. Whether the expression only includes AND (&&) logical operator,
default `true`.
    3. Whether the expression has NOT (!) operator, default `false`.
4. Store if turning on MV on the proxy to eliminate reading from
paramtable for every search request.
5. Renames to MV.

## Rebuttals
1. Did not write in `ExtractInfoPlanNodeVisitor` since the new scalar
framework was introduced and this part might be removed in the future.
2. Currently only interested in `==` and `in` expression, `string` data
type, anything else is a bonus.
3. Leave handling expressions like `F == A || F == A` for future works
of the optimizer.

## Detailed MV Info

![image](https://github.com/milvus-io/milvus/assets/6563846/b27c08a0-9fd3-4474-8897-30a3d6d6b36f)

Signed-off-by: Patrick Weizhi Xu <weizhi.xu@zilliz.com>
2024-03-21 11:19:07 +08:00
groot c81909bfab
enhance: Support MinIO TLS connection (#31311)
issue: https://github.com/milvus-io/milvus/issues/30709
pr: #31292

Signed-off-by: yhmo <yihua.mo@zilliz.com>
Co-authored-by: Chen Rao <chenrao317328@163.com>
2024-03-21 11:15:20 +08:00
yihao.dai c725f95885
enhance: Enhance the load balancing of import tasks (#31423)
Optimize the load balancing of import tasks.

issue: https://github.com/milvus-io/milvus/issues/31232

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-03-21 10:21:06 +08:00
congqixia 3f7c77481d
fix: Use correct ticker for gc work defer (#31456)
See also #31362

Fix the second defer using same ticker var.

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-03-20 22:17:06 +08:00
congqixia a647b84f3e
enhance: Add AllPartitionsID const to replace InvalidPartitionID (#31438)
"-1" as `InvalidPartitionID` previously used as All partition place
holder in delete cases. It's confusing and hard to maintain when a const
var has more than one meaning.

This PR add `AllPartitionsID` to replace these usages in delete
scenarios.

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-03-20 19:01:05 +08:00
aoiasd 0c153a5820
enhance: Rename update segment operator (#31121)
Signed-off-by: aoiasd <zhicheng.yue@zilliz.com>
2024-03-20 17:53:14 +08:00
cai.zhang 4dca57535f
fix: Fix bug for get segment index state (#31427)
issue: #31361

Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
2024-03-20 14:45:10 +08:00
Jiquan Long dc2cdbe387
enhance: add more metrics (#31271)
/kind improvement
fix: #31272 

This pr add more metrics, which are:
- Slow query count, which the duration considered as slow can be
configurable;
- Number of deleted entities;
- Number of entities imported;
- Number of entities per collection;
- Number of loaded entities per collection;
- Number of indexed entities;
- Number of indexed entities, per collection, per index and whether it's
a vetor index;
- Quota states (LongTimeTickDelay, MemoryExhuasted, DiskQuotaExhuasted)
per database;

---------

Signed-off-by: longjiquan <jiquan.long@zilliz.com>
2024-03-19 15:23:06 +08:00
congqixia 16c661c722
enhance: Use different interval for gc scan (#31363)
See also #31362

This PR make datacoord garbage collection scan operation using differet
interval than other opeartion.

This interval is a newly added param item, which default value is 7*24
hours.

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-03-19 11:27:06 +08:00
XuanYang-cn 0066c016b6
enhance: Skip submit empty l0 tasks in DC (#31280)
Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2024-03-19 10:13:14 +08:00
Xiaofan 8c43c5b6cb
fix: get compaction failure when datanode is actually alive (#31353)
didn't mark the compact as failure if it's simply an rpc error when
GetCompactionPlansResults
see #31352

Signed-off-by: xiaofanluan <xiaofan.luan@zilliz.com>
2024-03-18 10:01:36 -07:00
Bingyi Sun bdc70dfc6a
feat: Add global mmap enable configuration (#31267)
https://github.com/milvus-io/milvus/issues/31279

Signed-off-by: sunby <sunbingyi1992@gmail.com>
2024-03-18 15:17:10 +08:00
yihao.dai 776709e5ff
fix: Fix binlog import (#31310)
Fix binlog import functionality by removing the existing check and
refining the size retrieval process.

issue: https://github.com/milvus-io/milvus/issues/31221,
https://github.com/milvus-io/milvus/issues/28521

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-03-17 20:59:04 +08:00
cai.zhang 4871786a7b
enhance: When describing an index, fetch the index info in batches (#31238)
issue: #29313

Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
2024-03-15 16:37:09 +08:00
yihao.dai c408a32db6
feat: Add disk quota checks for import V2 (#31131)
Return quota error when the files to be imported exceed the disk quota.

issue: https://github.com/milvus-io/milvus/issues/28521

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-03-15 14:43:03 +08:00
yihao.dai 2b035ba2d4
enhance: Allow import tasks to retry for more errors (#31268)
Allow import tasks to retry for a wider range of errors, including all
gRPC errors and unexpected status codes from Milvus.

issue: https://github.com/milvus-io/milvus/issues/31227,
https://github.com/milvus-io/milvus/issues/28521

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-03-15 11:05:04 +08:00
yihao.dai 811316d2ba
fix: Fix binlog import and refine error reporting (#31241)
1. Fix binlog import with partition key.
2. Refine binlog import error reportins.
3. Avoid division by zero when retrieving import progress.

issue: https://github.com/milvus-io/milvus/issues/31221,
https://github.com/milvus-io/milvus/issues/28521

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-03-15 10:55:05 +08:00
jaime db79be3ae0
fix: ctx cancel should be the last step while stopping server (#31220)
issue: #31219

Signed-off-by: jaime <yun.zhang@zilliz.com>
2024-03-15 10:33:05 +08:00
XuanYang-cn a1386bae7f
fix: Skip to submit l0 tasks when scheduler full (#31270)
See also: #31242

Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2024-03-15 10:21:12 +08:00
yihao.dai 7d7ef388df
enhance: Remove adding import segments to the datanode (#31244)
With the presence of L0 segments, there's no longer a need to add import
segments to the datanode.

issue: https://github.com/milvus-io/milvus/issues/28521

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-03-15 06:53:03 +08:00
XuanYang-cn 1fafc72077
fix: Correct the last empty l0 views (#31198)
See also: #31191

---------

Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2024-03-14 10:31:04 +08:00
Buqian Zheng 3c80083f51
feat: [Sparse Float Vector] add sparse vector support to milvus components (#30630)
add sparse float vector support to different milvus components,
including proxy, data node to receive and write sparse float vectors to
binlog, query node to handle search requests, index node to build index
for sparse float column, etc.

https://github.com/milvus-io/milvus/issues/29419

---------

Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
2024-03-13 14:32:54 -07:00
yihao.dai b5c67948b7
enhance: Enhance and modify the return content of ImportV2 (#31192)
1. The Import APIs now provide detailed progress information for each
imported file, including details such as file name, file size, progress,
and more.
2. The APIs now return the collection name and the completion time.
3. Other modifications include changing jobID to jobId and other similar
adjustments.

issue: https://github.com/milvus-io/milvus/issues/28521

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-03-13 19:51:03 +08:00
yihao.dai 69e132e05b
Use logID instead of logPath for import segment (#31182)
Currently, the logPath in the querycoord should be replaced with logID.
This PR updates the import segment's logPath to logID.

issue: https://github.com/milvus-io/milvus/issues/31123,
https://github.com/milvus-io/milvus/issues/28885

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-03-12 10:13:03 +08:00
Bingyi Sun 5c0bb40549
fix: merge index params when creating index (#31127)
issue: https://github.com/milvus-io/milvus/issues/31102

---------

Signed-off-by: sunby <sunbingyi1992@gmail.com>
2024-03-11 17:31:03 +08:00
Bingyi Sun 425da78b38
fix: alter index request's index name can not be empty (#31128)
issue: https://github.com/milvus-io/milvus/issues/31138

---------

Signed-off-by: sunby <sunbingyi1992@gmail.com>
2024-03-11 13:05:02 +08:00
yihao.dai 8cb06acfed
feat: Replacing the current import API with the v2 implementation (#31046)
Replacing the current import API v1 implementation with the v2
implementation.

issue: https://github.com/milvus-io/milvus/issues/28521

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-03-10 12:23:02 +08:00
yihao.dai c411cb4a49
enhance: Prevent the backlog of channelCP update tasks, perform batch updates of channelCPs (#30941)
This PR includes the following adjustments:
1. To prevent channelCP update task backlog, only one task with the same
vchannel is retained in the updater. Additionally, the lastUpdateTime is
refreshed after the flowgraph submits the update task, rather than in
the callBack function.
2. Batch updates of multiple vchannel checkpoints are performed in the
UpdateChannelCheckpoint RPC (default batch size is 128). Additionally,
the lock for channelCPs in DataCoord meta has been switched from key
lock to global lock.
3. The concurrency of UpdateChannelCheckpoint RPCs in the datanode has
been reduced from 1000 to 10.

issue: https://github.com/milvus-io/milvus/issues/30004

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
Co-authored-by: jaime <yun.zhang@zilliz.com>
Co-authored-by: congqixia <congqi.xia@zilliz.com>
2024-03-07 20:39:02 +08:00
congqixia d81ba164c8
enhance: Add `ListIndexes` API from datacoord (#31104)
See also #31103

This PR add `listIndexes` API for datacoor server to list all indexes
for provided collection.
Comparing to the existing `DescribeIndex` API, the new one does NOT
check the segment index building progress to ease the burden when
invoking it

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-03-07 17:37:01 +08:00
congqixia 196f0c1e1d
fix: Skip invalid compaction plan (#31045)
See also #31044

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-03-06 21:36:59 +08:00
Bingyi Sun df7aafa3ec
fix: filter mmap key when checking index params (#31030)
issue: https://github.com/milvus-io/milvus/issues/31031

Signed-off-by: sunby <sunbingyi1992@gmail.com>
2024-03-06 16:03:00 +08:00
XuanYang-cn def72947c7
fix: Trigger l0 compaction when l0 views don't change (#30729)
Trigger l0 compaction when l0 views don't change

So that leftover l0 segments would be compacted in the end.

1. Refresh LevelZero plans in comactionPlanHandler, remove the meta
dependency
of compaction trigger v2
2. Add ForceTrigger method for CompactionView interface
3. rename mu to taskGuard
4. Add a new TriggerTypeLevelZeroViewIDLE
5. Add an idleTicker for compaction view manager

See also: #30098, #30556

Signed-off-by: yangxuan <xuan.yang@zilliz.com>

---------

Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2024-03-05 16:37:00 +08:00
zhagnlu b9775a1816
fix: add GetSegments optimization to avoid meta mutex competition (#31025)
#30835

Signed-off-by: luzhang <luzhang@zilliz.com>
Co-authored-by: luzhang <luzhang@zilliz.com>
2024-03-05 14:47:00 +08:00
congqixia 1936aa4caa
enhance: Check channel cp lag before generate compaction task (#30997)
See also #30996

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-03-05 13:39:01 +08:00
chyezh 8f7019468f
fix: starve lock caused by slow GetCompactionTo method when too much segments (#30963)
issue: #30823

Signed-off-by: chyezh <chyezh@outlook.com>
2024-03-05 10:04:59 +08:00
jaime 4b0c3dd377
enhance: index meta use independent rather than global meta lock (#30869)
issue: https://github.com/milvus-io/milvus/issues/30837

Signed-off-by: jaime <yun.zhang@zilliz.com>
2024-03-04 16:56:59 +08:00
cai.zhang f6ff2588cd
enhance: Optimize DescribeIndex to reduce lock contention (#30939)
issue: #29313 
issue: #30443

Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
2024-03-03 19:00:59 +08:00
yihao.dai a434d33e75
feat: Add import scheduler and manager (#29367)
This PR introduces novel managerial roles for importv2:
1. ImportMeta: To manage all the import tasks;
2. ImportScheduler: To process tasks and modify their states;
3. ImportChecker: To ascertain the completion of all tasks and instigate
relevant operations.

issue: https://github.com/milvus-io/milvus/issues/28521

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-03-01 18:31:02 +08:00
congqixia 6387403639
fix: Prevent clone when selecting segments from meta (#30928)
See also #30538

Previously the `SelectSegments` changed to clone all return value
preventing possible update to returned info.

Since meta is implemented following COW rules, this shall not happen and
any update on segment shall have copy before it.

This PR:
- Remove clone for read-only Get segment info
- Add Segment Operator abstraction for changing segment
- Implemnt COW for updating MaxRowNum

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-03-01 13:59:02 +08:00
aoiasd 2180e2cfae
fix: wrong segment binlog path cause load segment failed #30726 (#30959)
DataCoord GetSegmentInfo should return binlog info with logpath instead
logid when segment merge child segment's
binlog.
relate: https://github.com/milvus-io/milvus/issues/30366
https://github.com/milvus-io/milvus/issues/30165
https://github.com/milvus-io/milvus/issues/30550

Signed-off-by: aoiasd <zhicheng.yue@zilliz.com>
2024-03-01 13:09:01 +08:00
XuanYang-cn 2867f50fcc
fix: Clear DN unkown compaction tasks (#30850)
If DC restarted,  those unkonwn compaction tasks
will never get call back in DN, so that the segments in the compaction
task will be locked, unable to sync and compaction again, blocking cp
advance and compaction executing.

See also: #30137

---------

Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2024-03-01 11:31:00 +08:00
chyezh 0c7474d7e8
enhance: add graceful stop timeout to avoid node stop hang under extreme cases (#30317)
1. add coordinator graceful stop timeout to 5s
2. change the order of datacoord component while stop
3. change querynode grace stop timeout to 900s, and we should
potentially change this to 600s when graceful stop is smooth

issue: #30310
also see pr: #30306

---------

Signed-off-by: chyezh <chyezh@outlook.com>
2024-02-29 17:01:50 +08:00
Bingyi Sun 816ed671aa
fix: alter_index should return error if index not found (#30786)
issue: https://github.com/milvus-io/milvus/issues/30932

Signed-off-by: sunby <sunbingyi1992@gmail.com>
2024-02-29 15:55:01 +08:00
XuanYang-cn cdc5ce5d6f
fix: Donot set LogPath when executing compaction (#30537)
Compaction would copy logPaths from comapctFrom segA to compactTo segB,
and previous code would copy the logPath directly, causing there're
full-logPaths-of-segA in compactTo segB's meta. So, for the next
compaction of segB, if segA has been GCed, Download would report error
"The sperified key not found".

This PR makes sure compactTo segment's meta contains logID only. And
this PR also refines CompleteComapctionMutation, increasing some
readability and merge two methods into one.

See also: #30496

Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2024-02-28 19:03:01 +08:00