1647bebaf4
Count growing segment number to estimate cpu usage ( #17820 ) ( #18027 )
...
Signed-off-by: zhenshan.cao <zhenshan.cao@zilliz.com>
2022-07-04 15:10:20 +08:00
6b79d4f7bc
Log segment type as possible in QueryNode ( #17846 )
...
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2022-06-28 10:16:17 +08:00
16c3aedc15
refine complie configuration ( #17502 )
...
Signed-off-by: Enwei Jiao <enwei.jiao@zilliz.com>
2022-06-24 21:12:15 +08:00
5f44e45480
Add metrics on QueryNode and Proxy ( #17328 )
...
Signed-off-by: zhenshan.cao <zhenshan.cao@zilliz.com>
2022-06-02 16:06:03 +08:00
63a31ccb38
Remove error return value when removeSegment ( #17310 )
...
Signed-off-by: xige-16 <xi.ge@zilliz.com>
2022-06-01 13:18:02 +08:00
f55fe6be3c
Merge historical replica and streaming replica in QueryNode ( #17217 )
...
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2022-05-31 13:42:03 +08:00
a4ea2fb18a
Handle errors in DataNode and QueryNode flow graph ( #17096 )
...
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2022-05-24 21:11:59 +08:00
08a4dff41d
Add logic for search merging and a simple task scheduler for read tasks ( #17022 )
...
Signed-off-by: zhenshan.cao <zhenshan.cao@zilliz.com>
Co-authored-by: cai.zhang <cai.zhang@zilliz.com>
Co-authored-by: bigsheeper <yihao.dai@zilliz.com>
Co-authored-by: cai.zhang <cai.zhang@zilliz.com>
Co-authored-by: bigsheeper <yihao.dai@zilliz.com>
2022-05-23 16:41:58 +08:00
94f0951f0d
Fix query lock logic in query shard ( #17034 )
...
Previously query shard locks the querylock in collectionReplica before any search/query
The lock range is too large and easy to cause dead lock
This PR makes following changes:
- Rename collectionReplica to metaReplica which is more reasonable
- Make release collection operation cancels waiting search/query request
- Reduce the queryLock to collection level
- Add some unit tests for timeout & released case
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2022-05-16 18:23:55 +08:00
zhenshan.cao
bigsheeper
Enwei Jiao
xige-16
congqixia