issue: #29261
This PR Add restful api for devops to execute rolling upgrade, including
suspend/resume balance and manual transfer segments/channels.
---------
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
issue: #31480#31481
release duplicate l0 segment task, which execute on old delegator may
cause segment lack, and execute on new delegator may break new
delegator's leader view.
This PR skip release duplicate l0 segment by segment_checker, cause l0
segment will be released with unsub channel
---------
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
See also #28491#31240
When colleciton number is large, querycoord saves collection target one
by one, which is slow and may block querycoord exits.
In local run, 500 collections scenario may lead to about 40 seconds
saving collection targets.
This PR changes the `SaveCollectionTarget` interface into batch one and
organizes the collection in 16 per bundle batches to accelerate this
procedure.
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
issue: #30816
pr #31319 introduce the logic that segment checker need to load level
zero segment which only exist in current target.
This PR fix load segment task promote failed when segment only belongs
to current target
---------
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
issue: #28491
after querycoord restart, it will pull a new target, which include
channel and segment list. when segments loaded on querynode has reached
the target, the collection could provide search/query. but if segment
list changes by time, ater querycoord pull a new target, it will takes a
few minutes to catch up the target's segment distribution. and before
that, query/search will fail due to lack of segments.
This PR save the current loaded target to meta storein querycoord's stop
progress, and recover it when query coord starts, to speed up the target
recovery time.
---------
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
issue: #31222
grpcclient's `call` func return a unrecoverable error, then the caller's
retry policy also breaks due to this unrecoverable error.
This PR introduce `retry.Handle`, the new func use `func() (bool,
error)` as input parameters, which return `shouldRetry` directly, to
avoid grpcclient return a unrecoverable error
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
issue: #30816
cause balance channel will stuck until leader view catch up the current
target, then start to unsub the old delegator. which make sure that the
new delegator can provide search before release old delegator. but
another logic in segment_checker skip loading segment during balance
channel. so during balance channel, if query node crash, new delegator
can't catch up target forever, then stuck forever.
This PR remove the rule that skip loading segment during balance channel
to avoid the logic dead lock here.
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
issue: #30983#30982
cause balancer call wrong interface to get segment/channel list in
replica, then got a wrong average segment/channel number, which make
each node have less segment/channel than average, and the balance won't
be trigger in multi replica case.
This PR fix that balance segment/channel won't be trigger on multi
replicas
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
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>
Now the integration tests have lots of time during tear down phase when
there are some segments or channel left in querynode(s). Querynode
cannot quick as soon as test finishes since graceful shut down logic
need querynode has no segment or channel in it. Auto releasing loaded
collection could reduce lots of running time for integration test.
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
- Set log level to INFO for integration test
- Improve hello_milvus integration test runtime and error check
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
* When the collection is created using an SDK and includes array and
JSON datatypes in the schema, data can be inserted using the RESTful
API.
* When the collection is created using the RESTful API and includes JSON
and array datatypes in dynamic fields, data can also be inserted using
the RESTful API.
Signed-off-by: zhuwenxing <wenxing.zhu@zilliz.com>
issue: #31063
Increase the graceful time to ensure that bounded query requests can be
executed normally in the event of RootCoord failure.
Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
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>
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>
issue: #30715
- Bug: Set nil struct pointer to describe nil interface.
Panic with segment violation when calling method on this nil struct
pointer.
Signed-off-by: chyezh <chyezh@outlook.com>
This PR mainly improve two items:
1. Target observer should refresh loading status during init time. An
uninitialized loading status blocks search/query. Currently, the target
observer refreshes every 10 seconds, i.e. we'd need to wait for 10s for
no reason. That's also the reason why we constantly see false log
"collection unloaded" upon mixcoord restarts.
2. Delete session when service is stopped. So that the new service
doesn't need to wait for the previous session to expire (~10s).
Item 1 is the major improvement of this PR, which should speed up init
time by 10s.
Item 2 is not a big concern in most cases as coordinators usually shut
down after stop(). In those cases, coordinator restart triggers serverID
change which further triggers an existing logic that deletes expired
session. This PR only fixes rare cases where serverID doesn't change.
integration test:
`go test -tags dynamic -v -coverprofile=profile.out -covermode=atomic
tests/integration/coordrecovery/coord_recovery_test.go -timeout=20m`
Performance after the change:
Average init time of coordinators: 10s
Hardware: M2 Pro
Test setup: 1000 collections with 1000 rows (dim=128) per collection.
issue: #29409
Signed-off-by: yiwangdr <yiwangdr@gmail.com>
This pr decoups importing segment from flush process by:
1. Exclude the importing segment from the flush policy, this approch
avoids notifying the datanode to flush the importing segment, which may
not exist.
2. When RootCoord call Flush, DataCoord directly set the importing
segment state to `Flushed`.
issue: https://github.com/milvus-io/milvus/issues/30359
---------
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
fix query result verification:
changed the query expression and adopted a more lenient validation
method to address the issue of not being able to guarantee the retrieval
of specific IDs due to frequent deletion operation
Signed-off-by: zhuwenxing <wenxing.zhu@zilliz.com>
issue: #29087
RBAC cases fail a lot.
1. some cases are out of date, for example, the default value of db_name
has changed from "default" to "" in some apis
2. add time sleep after the action of grant or revoke, for it costs time
to take effect
Signed-off-by: nico <cheng.yuan@zilliz.com>
add freshness checker
insert/upsert --> query: Get the time when it can be queried
delete --> query: Get the time when it can not be queried
Signed-off-by: zhuwenxing <wenxing.zhu@zilliz.com>
1. modify test case: test_search_repeatedly_ivf_index_different_limit
2. update pymilvus version from 2.4.0rc19 to 2.4.0rc24
3. Before, insert will return a pk list. In the latest milvus client,
insert will return a number that is inserted successfully
4. In the latest milvus client, flush and num_entities have been removed
5. Before, the default consistency level of a new collection is strong.
In the latest milvus client, it becomes bounded. So related cases have
been modified correspondingly, or immediate search after insert will
return no results.
6. In the latest pymilvus, new data type FLOAT16_VECTOR and
BFLOAT16_VECTOR have been added.
Signed-off-by: nico <cheng.yuan@zilliz.com>
don't store logPath in meta to reduce memory, when service get
segmentinfo, generate logpath from logid.
#28885
Signed-off-by: lixinguo <xinguo.li@zilliz.com>
Co-authored-by: lixinguo <xinguo.li@zilliz.com>
currently integration test may timeout if any case run time is above 3
minutes. This duration was hard coded.
This PR change this duration into a customized parameter and could be
passed via test running commands.
---------
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
pr: https://github.com/milvus-io/milvus/pull/29634
After verifying that the problem is fixed, it is necessary to run this
case every night for regression.
Signed-off-by: elstic <hao.wang@zilliz.com>
This pull request simplifies the integration test for cross-cluster
routing by reusing `integration.MiniClusterSuite`, instead of defining
custom Milvus clients, servers, and etcd client.
issue: https://github.com/milvus-io/milvus/issues/29874
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
See also: #29113
Add a new utitliy function in `pkg/util/typetuil` to pre-allocate field
data slice capacity acoording to search limit. This shall avoid copying
the data during `AppendFieldData` when previous slice is out of space.
And shall also save CPU time during high paylog.
---------
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
Once a role is granted to a user, the user should automatically possess
the privilege information associated with that role.
issue: #29710
Signed-off-by: zhenshan.cao <zhenshan.cao@zilliz.com>
Memory management is broken due caused by chunkmanager not getting
cleaned up.
Fixing the following error:
```
SIGSEGV: segmentation violation
PC=0x10fc4aa0c m=6 sigcode=2
```
relate: https://github.com/milvus-io/milvus/issues/29507
Signed-off-by: yiwangdr <yiwangdr@gmail.com>
issue: #17614
update some new api: load_state, num_shards, get_server_type
remove some old api: collection.create_alias, drop_alias
Signed-off-by: nico <cheng.yuan@zilliz.com>
1. modify test cases about output a non-existed field. issue: #28465
2. add cases about creating collection using different language
3. add test case about inserting a string value to a pk(int64) field
4. add test case about search with expression using double quotes.
issue: #28365
Signed-off-by: nico <cheng.yuan@zilliz.com>
1. Drop a non-existing partition will not raise error now. #28722
2. Search will not raise error now when some row has less elements than
expression index. #28293
Signed-off-by: nico <cheng.yuan@zilliz.com>
test: There are too many test cases for bulkinsert+partition_key. Each
case creates 10 bulkinsert tasks to import a file with 100~200 rows. The
default num_partitions is 64 for partition_key. So, each task will
generate 64 tiny segments. There are 10 cases, each case 10 tasks, each
task 64 tiny segment, totally there are 6400 tiny segments generated.
And all these segment row count is less than 1024, no need to build
index, and take part in compaction. There will be lots of compaction
tasks generated. It costs too much time to process these compaction
tasks. Eventually, some cases are timeout after waiting 5 minutes for
their segments to be ready and cases fail.
Specifying the num_partitions to a small value can avoid this problem.
```
[2023-11-21T03:41:16.187Z] testcases/test_bulk_insert.py::TestBulkInsert::test_partition_key_on_json_file[int_scalar-True-True] PASSED [ 54%]
[2023-11-21T03:41:42.796Z] testcases/test_bulk_insert.py::TestBulkInsert::test_partition_key_on_json_file[int_scalar-False-True] PASSED [ 57%]
[2023-11-21T03:42:04.694Z] testcases/test_bulk_insert.py::TestBulkInsert::test_partition_key_on_json_file[string_scalar-True-True] PASSED [ 60%]
[2023-11-21T03:42:31.205Z] testcases/test_bulk_insert.py::TestBulkInsert::test_partition_key_on_json_file[string_scalar-False-True] PASSED [ 63%]
[2023-11-21T03:43:38.876Z] testcases/test_bulk_insert.py::TestBulkInsert::test_partition_key_on_multi_numpy_files[10-150-13-True] XPASS [ 66%]
[2023-11-21T03:49:00.357Z] testcases/test_bulk_insert.py::TestBulkInsert::test_partition_key_on_multi_numpy_files[10-150-13-False] XFAIL [ 69%]
[2023-11-21T03:53:51.811Z] testcases/test_bulk_insert.py::TestBulkInsert::test_partition_key_on_csv_file[int_scalar-True] FAILED [ 72%]
[2023-11-21T03:58:58.283Z] testcases/test_bulk_insert.py::TestBulkInsert::test_partition_key_on_csv_file[int_scalar-False] FAILED [ 75%]
[2023-11-21T04:02:04.696Z] testcases/test_bulk_insert.py::TestBulkInsert::test_partition_key_on_csv_file[string_scalar-True] PASSED [ 78%]
[2023-11-21T04:02:26.608Z] testcases/test_bulk_insert.py::TestBulkInsert::test_partition_key_on_csv_file[string_scalar-False] PASSED [ 81%]
```
Signed-off-by: yhmo <yihua.mo@zilliz.com>
zhuwenxing
binbin
SimFG
congqixia
wei liu
nico
yihao.dai
yanliang567
cai.zhang
sammy.huang
elstic
sre-ci-robot
Bingyi Sun
XuanYang-cn
Xiaofan
jaime
zhagnlu
chyezh
Jiquan Long
yiwangdr
aoiasd
zhenshan.cao
yah01
smellthemoon
wayblink
Xu Tong
xige-16
groot