yihao.dai comments

Results 263 comments of


                                            yihao.dai

[Bug]: flush timeout after milvus recovered from pulsar pod kill chaos test

> reproduced in `master-20240709-eeb03a0d-amd64` datanode chaos test > > failed job: https://qa-jenkins.milvus.io/blue/organizations/jenkins/chaos-test-kafka-cron/detail/chaos-test-kafka-cron/15588/pipeline > > log:[artifacts-datanode-pod-kill-15588-server-logs.tar.gz](https://github.com/user-attachments/files/16154356/artifacts-datanode-pod-kill-15588-server-logs.tar.gz) dispatcher merged by mistake, the insert data consumed was incorrectly skipped: ![image](https://github.com/milvus-io/milvus/assets/42060877/86c2461e-4bd0-450f-bcb3-b9fc7022655e)

[Bug]: flush timeout after milvus recovered from pulsar pod kill chaos test

/assign @zhuwenxing should be fixed

[Bug]: flush timeout after milvus recovered from pulsar pod kill chaos test

/unassign

[Bug]: [benchmark][cluster] DQL request raises error `channel distribution is not servicea ble: channel not available` in concurrent DML & DQL scene

/assign

[Bug]: [benchmark][cluster] DQL request raises error `channel distribution is not servicea ble: channel not available` in concurrent DML & DQL scene

Same as https://github.com/milvus-io/milvus/issues/45301:

fix: listImport and getImportProgress should follow import access

/lgtm

fix: [2.6]listImport and getImportProgress should follow import access

/refresh-label

[Bug]: During the rolling upgrade process, the datanode experiences a sudden surge in memory usage, leading to OOM kills, and the querynode also continues to crash

This PR (https://github.com/milvus-io/milvus/pull/34278) accelerates the subscription speed of the dispatcher. However, the subscription becomes too fast for the dispatcher to merge in time, causing the DataNode OOM. **Test with PR...

[Bug]: During the rolling upgrade process, the datanode experiences a sudden surge in memory usage, leading to OOM kills, and the querynode also continues to crash

PRs https://github.com/milvus-io/milvus/pull/40596 and https://github.com/milvus-io/milvus/pull/40854 should be highly beneficial for this case. Please help to verify. @zhuwenxing

[Bug]: During the rolling upgrade process, the datanode experiences a sudden surge in memory usage, leading to OOM kills, and the querynode also continues to crash

/assign @zhuwenxing