848 Commits

Author SHA1 Message Date
yihao.dai
917a4d74f3
fix: Use channel cp as the dml&start position for import segments (#30107) (#30133)
This PR discontinuing the subscription to the mq and, instead, employing
the channel checkpoint as the DML and starting position for the import
segments.

issue: https://github.com/milvus-io/milvus/issues/30106

pr: https://github.com/milvus-io/milvus/pull/30107

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-01-22 13:32:55 +08:00
congqixia
1dbc2ab8ee
enhance: [Cherry-pick] make compactor use actual buffer size to decide when to sync(#29945) (#29971)
Cherry-pick from master
pr: #29945
See also: #29657

Datanode Compactor use estimated row number from schema to decide when
to sync the batch of data when executing compaction. This est value
could go way from actual size when the schema contains variable field(
say VarChar, JSON, etc.)

This PR make compactor able to check the actual buffer data size and
make it possible to sync when buffer is actually beyond max binglog
size.

---------

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-01-16 12:22:52 +08:00
SimFG
d573f0ec1a
fix: [2.3] the delete msg disorder issue (#29917)
/kind improvement
pr: #29915

Signed-off-by: SimFG <bang.fu@zilliz.com>
2024-01-12 18:04:50 +08:00
congqixia
c56622dea7
enhance: move confusing warning log to error branch (#29891)
`flushInsertData` & `flushDeleteData` prints WARNING log even there is
no error returned. So move error branch into if block.

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-01-12 15:50:52 +08:00
XuanYang-cn
1128b1dd67
fix: [cherry-pick]Save lite WatchInfo into etcd in DataNode (#29751)
See also: #29689
pr: #29687

Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2024-01-10 20:48:50 +08:00
congqixia
cf228c2f1c
fix: Print number of pks instead of delete pk val (#29653)
See also #29445

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-01-04 10:02:46 +08:00
MrPresent-Han
757834602a
enhance: add param for bloomfilter(#29388) (#29614)
related: https://github.com/milvus-io/milvus/issues/29388
pr: https://github.com/milvus-io/milvus/pull/29490

Signed-off-by: MrPresent-Han <chun.han@zilliz.com>
2024-01-02 18:58:47 +08:00
yah01
51cab791cf
fix: missing to support compact for Array type (#29505) (#29504)
the array type can't be compacted, the system could continue with the
inserted segments, but these segments can be never compacted

fix #29503
pr: #29505

---------

Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-12-27 08:15:51 +08:00
SimFG
74e72ce27e
enhance: [2.3] Support to get the param value in the runtime (#29298)
pr: #29297
/kind improvement

Signed-off-by: SimFG <bang.fu@zilliz.com>
2023-12-21 20:36:43 +08:00
Xiaofan
8e13199da2
fix: frequent flush cause minio rate limit (#28625)
related to #28549
pr: #28626

1. avoid duplicated sync segments under syncing states
2. add jitter to avoid sync segments at the same time

Signed-off-by: xiaofanluan <xiaofan.luan@zilliz.com>
2023-12-20 15:02:55 +08:00
congqixia
9a80dc666b
fix: flushTs is never reset in channelMeta (#29244)
See also #29156
FlushTs need to to be reset to MaxUint64 after channel checkpoint is
after this timestamp. Otherwise, the segment will be shattered and flush
queue will be filled with tasks

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2023-12-16 14:00:38 +08:00
XuanYang-cn
7b0599765f
fix: [cherry-pick]Skip updating checkpoint after dropcollection (#29221)
pr: #29220

Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2023-12-15 16:04:45 +08:00
congqixia
a108bf7bc1
enhance: improve datanode channel checkpoint source log (#29180)
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2023-12-14 14:34:38 +08:00
wayblink
e49860cb80
feat: Introduce channelCheckpointUpdater to reduce goroutine in ttNode (#29107)
pr: #28570

Signed-off-by: wayblink <anyang.wang@zilliz.com>
2023-12-12 13:48:42 +08:00
MrPresent-Han
5f4ac437b2
enhance: [Cherry-pick] Moving etcd client into session (#27069) (#28996)
relate: #26694
pr: https://github.com/milvus-io/milvus/pull/27069

Signed-off-by: Filip Haltmayer <filip.haltmayer@zilliz.com>
Signed-off-by: MrPresent-Han <chun.han@zilliz.com>
Co-authored-by: Filip Haltmayer <81822489+filip-halt@users.noreply.github.com>
2023-12-07 16:22:34 +08:00
congqixia
2873be9264
fix: [2.3] Reject compaction task with growing segments (#28927)
See also #28924
The compaction task generated before datanode finish SaveBinlogPath grpc
call contains segments which are still in Growing state DataNode shall
verify each non-levelzero segments before submit compaction task to
executor

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2023-12-04 19:14:33 +08:00
yihao.dai
a78ea4fea0
fix: Check ErrSegmentNotFound in delete node (#28371) (#28638)
We have been check ErrSegmentNotFound in insert_buffer_node in datanode,
we should also check it in delete_node.

issue: https://github.com/milvus-io/milvus/issues/27145

pr: https://github.com/milvus-io/milvus/pull/28371

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2023-11-29 17:50:27 +08:00
jaime
9378f78218
enhance: Add logs for each step during service initialization (#28687)
/kind improvement
pr: #28624

Signed-off-by: jaime <yun.zhang@zilliz.com>
2023-11-27 17:54:26 +08:00
congqixia
6512b12fba
enhance: [cherry-pick] Make etcd kv request timeout configurable (#28661) (#28701)
Cherry-pick from master
pr: #28661
See also #28660
This pr add request timeout config item for etcd kv request timeout
 Sync the default timeout value to same value for etcdKV & tikv config

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2023-11-24 21:16:26 +08:00
smellthemoon
288844e3cf
enhance: Reduce the goroutine in flowgraph to 2 (#28233) (#28545)
each node in flow graph alloc a goroutine, but it is actually executed
sequentially and can be placed in one goroutine. InputNode will consume
msg form msgstream, alloc one goroutine.
issue: #24826 
pr: #28233

Signed-off-by: lixinguo <xinguo.li@zilliz.com>
Co-authored-by: lixinguo <xinguo.li@zilliz.com>
2023-11-24 15:00:26 +08:00
smellthemoon
d2ebbe2317
enhance: create goroutine only once(#28594) (#28609)
create goroutine only once when getOrCreateMergedTimeTickerSender
pr: #28594

Signed-off-by: lixinguo <xinguo.li@zilliz.com>
Co-authored-by: lixinguo <xinguo.li@zilliz.com>
2023-11-22 10:26:28 +08:00
smellthemoon
d724b07037
enhance: Use single instance for mergedTimeTickerSender (#27730) (#28546)
use single instance for mergedTimeTickerSender
issue: https://github.com/milvus-io/milvus/issues/24826
pr: https://github.com/milvus-io/milvus/pull/27730

Signed-off-by: lixinguo <xinguo.li@zilliz.com>
Co-authored-by: lixinguo <xinguo.li@zilliz.com>
2023-11-21 16:04:23 +08:00
congqixia
fce0284881
[2.3] Refine datanode Timetick Sender (#28393) (#28430)
cherry pick from master
pr: #28393
- Use explicit lifetime control methods: `Start` and `Stop`
- Allow control retry option
- Make sure tt sender worker exit after `Stop` return

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2023-11-15 10:08:26 +08:00
yah01
e51ceaae3a
Not convert legacy error code to new merr (#28232) (#28274)
Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-11-08 19:18:20 +08:00
groot
29e66ed46b
Fix bulkinsert bug that segments are compacted after import (#28227)
Signed-off-by: yhmo <yihua.mo@zilliz.com>
2023-11-08 10:18:20 +08:00
SimFG
598788e6b8
Delay the cancellation of ctx when stopping the node (#28249)
Signed-off-by: SimFG <bang.fu@zilliz.com>
2023-11-08 01:46:20 +08:00
yah01
d10a82dba4
Fix getting incorrect CPU num (#28178)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-11-07 11:52:22 +08:00
yihao.dai
5fae32f77e
Use merr to prevent datanode panic (#28122)
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2023-11-04 00:58:21 +08:00
yah01
f79c7370f4
Fix panic while flushing dropped/compacted segment (#27927)
Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-10-25 22:10:13 +08:00
yihao.dai
b9d5ef3599
Fix datanode ttNode goroutine leak (#27878)
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2023-10-24 18:22:10 +08:00
aoiasd
9091a27832
Add meta cache to datanode for L0 Delta (#27768)
Signed-off-by: aoiasd <zhicheng.yue@zilliz.com>
2023-10-23 19:42:10 +08:00
SimFG
9b0ecbdca7
Support to replicate the mq message (#27240)
Signed-off-by: SimFG <bang.fu@zilliz.com>
2023-10-20 14:26:09 +08:00
smellthemoon
4b0ec156b3
Set channel work pool size in datanode (#27728)
Signed-off-by: lixinguo <xinguo.li@zilliz.com>
Co-authored-by: lixinguo <xinguo.li@zilliz.com>
2023-10-19 08:28:08 +08:00
XuanYang-cn
7358c3527b
Add iterators (#27643)
See also: #27606

Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2023-10-18 19:34:08 +08:00
jaime
e386a62fae
Remove recollect segment stats during starting datacoord (#27410)
Signed-off-by: jaime <yun.zhang@zilliz.com>
2023-10-16 10:26:09 +08:00
jaime
ec1fe3549e
Add a stop hook to clean session (#27564)
Signed-off-by: jaime <yun.zhang@zilliz.com>
2023-10-16 10:24:10 +08:00
congqixia
82b2edc4bd
Replace manual composed grpc call with Broker methods (#27676)
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2023-10-13 09:55:34 +08:00
yah01
be980fbc38
Refine state check (#27541)
Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-10-11 21:01:35 +08:00
aoiasd
ec830e1240
fix use same logidx in different field (#27492)
Signed-off-by: aoiasd <zhicheng.yue@zilliz.com>
2023-10-11 18:49:33 +08:00
congqixia
cbb350c552
Add broker for datanode grpc operations (#27631)
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2023-10-11 17:03:34 +08:00
yah01
2df9908c6a
Map old error code to new error (#27616)
Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-10-11 12:43:33 +08:00
jaime
86ed2fd51f
[skip e2e] Fix unstable ut (#27618)
Signed-off-by: jaime <yun.zhang@zilliz.com>
2023-10-11 07:21:32 +08:00
XuanYang-cn
294ff74ca5
Add new compaction types (#27608)
See also: #27606

Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2023-10-11 06:49:32 +08:00
XuanYang-cn
56c94cdfa7
Add channel manager in DataNode (#27308)
Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2023-10-08 21:37:33 +08:00
congqixia
5d558623fe
Add revive sub-lints and fix existing problems (#27495)
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2023-10-07 20:53:38 +08:00
congqixia
80eb5434d3
Fix var-naming caused by old PRs (#27501)
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2023-10-07 14:43:32 +08:00
yah01
8394b3a1ec
Block creating new error from status reason (#27426)
Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-10-07 11:29:32 +08:00
Jiquan Long
0f14d18201
Optimize the codec code of session (#27360)
Signed-off-by: longjiquan <jiquan.long@zilliz.com>
2023-10-01 10:33:30 +08:00
XuanYang-cn
5c5f9aa05e
Enhance newDataSyncService (#27277)
- Add flowgraph.Assemble assembles nodes in flowgraph.go
- remove fgCtx in newDataSyncService
- Add newServiceWithEtcdTickler func, reduce param numbers to 3
- Remove unnecessary params
  - config.maxQueueLength, config.maxParallelish

See also: #27207

Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2023-09-27 11:07:25 +08:00
yah01
6539a5ae2c
Refine DataCoord status (#27262)
Signed-off-by: yah01 <yah2er0ne@outlook.com>
2023-09-26 17:15:27 +08:00