[Bug]: certain VarChar can't be inserted

[pingren]

Is there an existing issue for this?

• [X] I have searched the existing issues

Environment

- Milvus version: v2.1.1
- Deployment mode: Standalone / Cluster (I tried both).
- SDK version: pymilvus2.1.1 (also tried @zilliz/[email protected] for Node.js)
- OS: 1.22.11-gke.400 on Google Cloud

It happens on deployments (Standalone or Cluster) by Helm on GKE. 
I can't reproduce on my local milvus by Docker Compose.

Current Behavior

Inserting "ann-ounce" will work. However, if I insert anything string including "announce", the request will hang for a while and fail.

Expected Behavior

No response

Steps To Reproduce

Python Script to reproduce:

from pymilvus import CollectionSchema, FieldSchema, DataType, Collection, connections

connections.connect(
  alias="default", 
  host='PUT YOUR HOST HERE', 
  port='19530'
)
vector = FieldSchema(
  name="vector", 
  dtype=DataType.FLOAT_VECTOR, 
  dim=1
)
id = FieldSchema(
  name="id", 
  dtype=DataType.VARCHAR, 
  is_primary=True,
  max_length=64
)
varchar = FieldSchema(
  name="varchar", 
  dtype=DataType.VARCHAR,
  max_length=1024
)
schema = CollectionSchema(
  fields=[vector, id, varchar], 
  description="Test varchar insert"
)

collection = Collection(
    name="test_varchar", 
    schema=schema, 
    using='default', 
    shards_num=2,
)

data = [
  [[0.3]],
  ['test-id'],
  ['ann-ounce'],
]
mr = collection.insert(data)
print('insert VARCHAR ann-ounce ok', mr)

data = [
  [[0.3]],
  ['test-id2'],
  ['announce'],
]

mr = collection.insert(data)
print('insert VARCHAR announce ok', mr)

Milvus Log

No response

Anything else?

No response

[yanliang567]

/assign @jingkl /unassign

[yanliang567]

@jingkl please help to reproduce it locally

[jingkl]

I used the Milvus 2.1.1 tag and the pymilvus==2.1.1 standalone service and executed the tests without errors. The test results are as follows

[yanliang567]

I used the Milvus 2.1.1 tag and the pymilvus==2.1.1 standalone service and executed the tests without errors. The test results are as follows

any chance to retry on AWS and Google

[soothing-rain]

@pingren can you provide logs of the Milvus server?

[pingren]

Yes. But my Milvus server is on production. I am too busy right now to set up a new environment.

Hopefully I could reproduce the issue in a dev server and provide logs in holidays next weeks.

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions. Rotten issues close after 30d of inactivity. Reopen the issue with /reopen.

[pingren]

I can't reproduce this one on a new environment either. I will close the issue.

A few days ago, our Milvus on production encountered another issue, and search service down about 12 hours. Unfortunately, all metrics and status are “health” on K8s.

I decide to switched to another vector search solution after all of these unexpected bugs.