Commit Graph

212 Commits (f12574aaf34262c899fda8e4fbc86ab97ac44ee1)

Author SHA1 Message Date
wei liu 74da53c027
fix update load percentage (#23054)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2023-03-30 10:48:23 +08:00
wei liu e2096965c7
fix leader view (#23038)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2023-03-29 14:06:02 +08:00
yah01 dc6d4b913a
Fix partitions may be not recovered with double load partitions (#23061)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-03-28 21:38:02 +08:00
yah01 ae0f467c02
Fix segment/channel may be re-loaded/subscribed (#22969)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-03-27 18:28:00 +08:00
yah01 081572d31c
Refactor QueryNode (#21625)
Signed-off-by: yah01 <yang.cen@zilliz.com>
Co-authored-by: Congqi Xia <congqi.xia@zilliz.com>
Co-authored-by: aoiasd <zhicheng.yue@zilliz.com>
2023-03-27 00:42:00 +08:00
XuanYang-cn 93bc805933
Enhance ID allocator in DataNode (#22905)
Signed-off-by: yangxuan <xuan.yang@zilliz.com>
2023-03-23 19:43:57 +08:00
yah01 2b81933d13
Refine logs of DistHandler (#22879)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-03-21 18:11:57 +08:00
yah01 68b9cabb87
Fix GetShardLeader returns old leader (#22887)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-03-21 16:57:57 +08:00
yihao.dai 1f718118e9
Dynamic load/release partitions (#22655)
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2023-03-20 14:55:57 +08:00
yah01 3d8f0156c7
Refine scheduler & executor of QueryCoord (#22761)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-03-16 17:43:55 +08:00
wei liu bb5088e605
fix unassign from rg (#22747)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2023-03-16 14:27:55 +08:00
SimFG b57e476089
Fix the nil point about the session (#22748)
Signed-off-by: SimFG <bang.fu@zilliz.com>
2023-03-14 20:07:54 +08:00
yah01 21ba8182ee
Refine task create errors (#22745)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-03-14 18:51:53 +08:00
yah01 1a4732bb19
Use new errors to handle load failures cache (#22672)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-03-10 17:15:54 +08:00
yah01 90a5aa6265
Refine errors, re-define error codes (#22501)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-03-09 15:47:52 +08:00
wei liu 11f1f4226a
support replica observer assign node (#22604)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2023-03-08 18:57:51 +08:00
congqixia f3e2d4a39a
Fix querynode stop meet unexpected GetComponentStates (#22590)
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2023-03-07 09:51:50 +08:00
yah01 d4fccdd135
Add reason for spawning task (#22549)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-03-06 10:13:50 +08:00
jaime d126f06946
Decouple mq module from internal proto definition (#22536)
Signed-off-by: jaime <yun.zhang@zilliz.com>
2023-03-04 23:21:50 +08:00
wei liu c162c6ecc0
fix assign node err (#22479)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2023-03-01 11:11:47 +08:00
wei liu d433e95ce0
fix transfer node meta err (#22420)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2023-02-27 19:29:47 +08:00
Enwei Jiao 697dedac7e
Use cockroachdb/errors to replace other error pkg (#22390)
Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>
2023-02-26 11:31:49 +08:00
zhenshan.cao e768437681
Correct usage of Timer and Ticker (#22228)
Signed-off-by: zhenshan.cao <zhenshan.cao@zilliz.com>
2023-02-23 18:59:45 +08:00
wei liu a9a263d5a8
fix assign node to replica in nodeUp (#22323)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2023-02-23 14:15:45 +08:00
wei liu c3e8ad3629
fix balance generate reduce task (#22236)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2023-02-21 19:06:27 +08:00
wei liu 87a4ddc7e2
fix rg e2e (#22187)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2023-02-16 10:48:34 +08:00
wei liu 7b4511b8f4
fix transfer node (#22120)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2023-02-14 16:16:34 +08:00
cai.zhang 66b3566ac1
Update component state to healthy after start (#22118)
Signed-off-by: cai.zhang <cai.zhang@zilliz.com>
2023-02-10 17:40:32 +08:00
wei liu e9684ddb5a
refine rg capacity behavior (#22089)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2023-02-09 18:46:31 +08:00
wei liu d085abbd56
fix load collection with rg (#22083)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2023-02-09 16:24:31 +08:00
yah01 b1f31da77a
Fix activate standby server ignores all errors (#22073)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-02-09 15:24:31 +08:00
Bingyi Sun 4005508880
Fix load collection blocks when memory is limited. (#21990)
Signed-off-by: sunby <bingyi.sun@zilliz.com>
Co-authored-by: sunby <bingyi.sun@zilliz.com>
2023-02-06 16:07:53 +08:00
wei liu 5da6aade02
fix default capacity of default rg (#21965)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2023-02-06 15:55:56 +08:00
Ten Thousand Leaves 74e3cc64fb
Add refresh path to proxy (#21933)
/kind improvement

Signed-off-by: Yuchen Gao <yuchen.gao@zilliz.com>
2023-02-02 19:29:51 +08:00
jaime e14f96a8e4
[skip e2e]Fix missed logger with information (#21859)
Signed-off-by: jaime <yun.zhang@zilliz.com>
2023-01-30 11:51:49 +08:00
cai.zhang bcae97b125
Register after start to prevent there are tow coordinator at the same time (#21641) (#21707)
Signed-off-by: cai.zhang <cai.zhang@zilliz.com>
2023-01-30 11:11:50 +08:00
wei liu 73c44d4b29
resource group impl (#21609)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2023-01-30 10:19:48 +08:00
Bingyi Sun b0b0929c6c
Change CreateAt to UpdatedAt when checking timeout (#21769)
Signed-off-by: sunby <bingyi.sun@zilliz.com>
Co-authored-by: sunby <bingyi.sun@zilliz.com>
2023-01-18 17:29:43 +08:00
Ten Thousand Leaves defb7660c8
Add refresh option for LoadCollection/LoadPartitions interfaces (#21776)
/kind improvement

Signed-off-by: Yuchen Gao <yuchen.gao@zilliz.com>
2023-01-18 16:41:44 +08:00
Bingyi Sun b91bb5a729
Fix load timeout after next target updates (#21759)
Signed-off-by: sunby <bingyi.sun@zilliz.com>
Co-authored-by: sunby <bingyi.sun@zilliz.com>
2023-01-17 15:15:46 +08:00
yah01 c8f89907b6
Fix current target may be updated to an invalid target (#21742)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-01-17 11:41:51 +08:00
congqixia 5f3d3dc4fc
generate-mockery for querycoordv2, querynode and rootcoort (#21714)
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2023-01-16 13:59:42 +08:00
yah01 837e3162d7
Fix ready notifiers leak when collection released (#21712)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-01-14 21:55:41 +08:00
yah01 32fb409e57
Fix may update the current target to an unavailable target when node down (#21698)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-01-13 17:11:41 +08:00
yah01 c66ce4aeba
Add warning comments (#21697)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-01-13 16:43:41 +08:00
Enwei Jiao 90d9e165d4
Fix some configs not shown (#21653)
Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>
2023-01-13 15:31:41 +08:00
congqixia 782f942b6f
Add MLogger.WithRateGroup for logger (#21703)
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2023-01-13 14:11:40 +08:00
wayblink 6a722396bd
Integration test framework (#21283)
Signed-off-by: wayblink <anyang.wang@zilliz.com>
2023-01-12 19:49:40 +08:00
Enwei Jiao fb42466c65
Use opentelemetry (#21509)
Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>
2023-01-12 16:09:39 +08:00
Enwei Jiao 9fccbbb92b
Remove factory in querycoord (#21659)
Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>
2023-01-12 15:59:39 +08:00
cai.zhang e127cf7b99
Reset indexpb for upgrade (#21620)
Signed-off-by: cai.zhang <cai.zhang@zilliz.com>
2023-01-11 14:35:40 +08:00
bigsheeper 2146af1fb2
Return insufficient memory error when load failed (#21574)
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2023-01-10 20:35:39 +08:00
yah01 3b41a6931d
Add QueryCoord OWNERS file (#21552)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-01-10 16:15:37 +08:00
MrPresent-Han 6fb3542f2a
enable auto balance paramter(#21504) (#21507)
Signed-off-by: MrPresent-Han <jamesharden11122@gmail.com>
2023-01-06 14:45:35 +08:00
cai.zhang aa203acfb3
Low IndexCoord weight (#21548)
Signed-off-by: cai.zhang <cai.zhang@zilliz.com>
2023-01-06 14:21:37 +08:00
cai.zhang e5f408dceb
Merge IndexCoord and DataCoord (#21267)
Signed-off-by: cai.zhang <cai.zhang@zilliz.com>
2023-01-04 19:37:36 +08:00
yah01 7b39873ae0
limit the frequency of GetMetrics() log (#21514)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2023-01-04 17:39:35 +08:00
Jiquan Long ff2a68e65a
Fix collection not exist when tried to do recovery (#21471)
Signed-off-by: longjiquan <jiquan.long@zilliz.com>
2023-01-04 16:37:35 +08:00
SimFG df6ccfa104
Fix wrong bool param in the `getDistribution` method (#21463)
Signed-off-by: SimFG <bang.fu@zilliz.com>
2022-12-30 14:15:31 +08:00
SimFG e6d2849c9a
Check whether the node is stopping when using `load balance` api (#21438)
Signed-off-by: SimFG <bang.fu@zilliz.com>
2022-12-29 15:47:31 +08:00
wayblink 0bcedbd2bf
quick fix queryCoordV2 active-standby (#21395)
Signed-off-by: wayblink <anyang.wang@zilliz.com>
2022-12-28 10:49:29 +08:00
SimFG 6a29a964df
Fix queryCoord panic during query node down (#21400)
Signed-off-by: SimFG <bang.fu@zilliz.com>
2022-12-28 10:17:30 +08:00
Xiaofan d16b7c3c2d
Add more logs for debugging load balance (#21389)
Signed-off-by: xiaofan-luan <xiaofan.luan@zilliz.com>
2022-12-27 15:57:30 +08:00
yah01 396a85c926
Fix the number of executing tasks may break the limit (#21318)
Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-12-20 20:33:27 +08:00
bigsheeper fc10c74005
Use channel-cp as seekPosition when FromDmlCPLoadDelete (#21110)
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2022-12-13 16:17:22 +08:00
smellthemoon b04eb17088
Fix load timeout (#21156)
Signed-off-by: lixinguo <xinguo.li@zilliz.com>

Signed-off-by: lixinguo <xinguo.li@zilliz.com>
Co-authored-by: lixinguo <xinguo.li@zilliz.com>
2022-12-12 18:43:23 +08:00
yah01 5ba1a94858
Fix observers may update current target to a unfinished next target (#21107)
Signed-off-by: yah01 <yang.cen@zilliz.com>

Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-12-12 15:15:22 +08:00
aoiasd de0ab9e2cf
Refactor showConfigurations to allow return global config rather than only return config of this component (#21063)
Signed-off-by: aoiasd <zhicheng.yue@zilliz.com>

Signed-off-by: aoiasd <zhicheng.yue@zilliz.com>
2022-12-09 14:31:21 +08:00
yah01 9ebaa10dec
Add more logs for GetShardLeaders (#21046)
Also increase the heartbeatAvailableInterval from 2.5s to 10s

Signed-off-by: yah01 <yang.cen@zilliz.com>

Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-12-08 19:09:18 +08:00
Bingyi Sun 27219ce9c0
Fix load progress bug (#21069)
issue: #21023

Signed-off-by: sunby <bingyi.sun@zilliz.com>

Signed-off-by: sunby <bingyi.sun@zilliz.com>
Co-authored-by: sunby <bingyi.sun@zilliz.com>
2022-12-08 18:41:20 +08:00
Enwei Jiao 89b810a4db
Refactor all params into ParamItem (#20987)
Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>

Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>
2022-12-07 18:01:19 +08:00
SimFG f8cff79804
Support the graceful stop for the query node (#20851)
Signed-off-by: SimFG <bang.fu@zilliz.com>

Signed-off-by: SimFG <bang.fu@zilliz.com>
2022-12-06 22:59:19 +08:00
yah01 0297ab1a46
Make progress if any channel/segment was loaded on node (#20775)
Signed-off-by: yah01 <yang.cen@zilliz.com>

Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-12-06 18:29:19 +08:00
yah01 11e4445ef7
Check whether segments are fully loaded while fetching shard leaders (#20991)
Signed-off-by: yah01 <yang.cen@zilliz.com>

Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-12-06 18:05:18 +08:00
smellthemoon c49f20ea94
Load should base on the accurate loaded number (#20999)
Signed-off-by: lixinguo <xinguo.li@zilliz.com>

Signed-off-by: lixinguo <xinguo.li@zilliz.com>
Co-authored-by: lixinguo <xinguo.li@zilliz.com>
2022-12-06 17:13:18 +08:00
yah01 aec347e591
Check last heartbeat response time while fetching shard leaders (#20968)
Signed-off-by: yah01 <yang.cen@zilliz.com>

Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-12-05 15:09:20 +08:00
yah01 ddd29ea6ab
Optimize scheduler, increase merge tasks probability (#20922)
Signed-off-by: yah01 <yang.cen@zilliz.com>

Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-12-01 16:25:16 +08:00
Enwei Jiao 2ecdb4ba4a
Etcd config source support TLS (#20874)
Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>

Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>
2022-11-30 18:23:15 +08:00
Jiquan Long 5a6a92d603
Refine log level (#20896)
Signed-off-by: longjiquan <jiquan.long@zilliz.com>

Signed-off-by: longjiquan <jiquan.long@zilliz.com>
2022-11-30 14:11:15 +08:00
yah01 060649b8aa
Refactor task scheduler and executor (#20828)
Make the performance able to scale out

Signed-off-by: yah01 <yang.cen@zilliz.com>

Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-11-30 13:57:15 +08:00
MrPresent-Han b021d7c59c
[skip e2e] add log for querynode load and remove segment(#20510) (#20810)
issue: #20510

Signed-off-by: MrPresent-Han <jamesharden11122@gmail.com>

Signed-off-by: MrPresent-Han <jamesharden11122@gmail.com>
2022-11-25 17:33:13 +08:00
wei liu 67403fcb3b
fix mannual balance with empty segment list (#20738)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2022-11-21 19:29:12 +08:00
smellthemoon 8283d32ac4
Only balance segement in targets (#20709)
Signed-off-by: lixinguo <xinguo.li@zilliz.com>

Signed-off-by: lixinguo <xinguo.li@zilliz.com>
Co-authored-by: lixinguo <xinguo.li@zilliz.com>
2022-11-21 16:21:11 +08:00
MrPresent-Han d44d50e735
fix getting query segment info error during the period of loading and unloading segments (#20549)
issue: #20281

Signed-off-by: MrPresent-Han <jamesharden11122@gmail.com>

Signed-off-by: MrPresent-Han <jamesharden11122@gmail.com>
2022-11-21 10:41:10 +08:00
Enwei Jiao c05b9ad539
Add event dispatcher for config (#20393)
Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>

Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>
2022-11-17 18:59:09 +08:00
yah01 cc371d6801
Task/Action won't finish util the RPC returned (#20669)
Signed-off-by: yah01 <yang.cen@zilliz.com>

Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-11-17 17:55:09 +08:00
Enwei Jiao 19524a5344
Fix nodeID mismatch at standalone mode (#20648)
Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>

Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>
2022-11-17 17:15:08 +08:00
wei liu 1177ed1a33
fix duplicate watch channel task (#20667)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2022-11-17 15:05:08 +08:00
yah01 31872f436c
Only balance segments in targets (#20635)
Signed-off-by: yah01 <yang.cen@zilliz.com>

Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-11-16 14:33:11 +08:00
yah01 2d806788fb
Add version for subscribed channel (#20585)
Signed-off-by: yah01 <yang.cen@zilliz.com>

Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-11-15 13:21:07 +08:00
yah01 d37ebe538b
Record load latency for loaded collection (#20538)
Signed-off-by: yah01 <yang.cen@zilliz.com>

Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-11-15 10:11:07 +08:00
smellthemoon 7325b3e1c3
Substitute traceid for msgid in rpc (#20450)
Signed-off-by: lixinguo <xinguo.li@zilliz.com>

Signed-off-by: lixinguo <xinguo.li@zilliz.com>
Co-authored-by: lixinguo <xinguo.li@zilliz.com>
2022-11-14 15:29:06 +08:00
wei liu 6a2e458f90
fix collection not exist cause query coord panic (#20553)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2022-11-14 14:51:06 +08:00
yah01 a82c235599
Scheduler should returns err if the queue is full (#20465)
Signed-off-by: yah01 <yang.cen@zilliz.com>

Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-11-11 17:55:05 +08:00
wei liu eaa5cfdcb5
fix leader observer sync logic (#20478)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2022-11-11 11:43:06 +08:00
wei liu 7537dbfa37
skip balance on loading collection (#20483)
Signed-off-by: Wei Liu <wei.liu@zilliz.com>

Signed-off-by: Wei Liu <wei.liu@zilliz.com>
2022-11-10 17:53:04 +08:00
yah01 c71c6378ff
Clear stale replicas (#20456)
Signed-off-by: yah01 <yang.cen@zilliz.com>

Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-11-10 17:01:03 +08:00
yah01 a68cec85af
Refine QueryCoordV2 metrics (#20461)
Signed-off-by: yah01 <yang.cen@zilliz.com>

Signed-off-by: yah01 <yang.cen@zilliz.com>
2022-11-10 15:01:04 +08:00
Jiquan Long 854fb6aad2
Fix querycoord metrics (#20429)
Signed-off-by: longjiquan <jiquan.long@zilliz.com>

Signed-off-by: longjiquan <jiquan.long@zilliz.com>
2022-11-09 17:45:04 +08:00
SimFG 9c6436d72d
Add time log for methods of starting the node (#20313)
Signed-off-by: SimFG <bang.fu@zilliz.com>

Signed-off-by: SimFG <bang.fu@zilliz.com>
2022-11-08 20:13:03 +08:00