[Bug]: CrashLoopBackOff DataNode During BulkImport

[akmalmasud96]

Is there an existing issue for this?

• I have searched the existing issues

Environment

- Milvus version:v2.5.1
- Deployment mode(standalone or cluster): Aws EKS Cluster
- MQ type(rocksmq, pulsar or kafka):    pulsar
- SDK version(e.g. pymilvus v2.0.0rc2): 2.5.2

Current Behavior

I have been importing the data into the milvus with parquet file, using bulk-import. However while importing, the data node is getting into the CrashLoopBackOff, and from the get_import_progress is showing the status for the files to be "Importing" from the last 5 hours.
also Is it not that if the task is not imported the duration after 15 minutes import task will be expired ( default )?

Expected Behavior

No response

Steps To Reproduce

resp = bulk_import(url="http://@localhost:19530",
                   collection_name = "collectionName",
                   partition_name = "partitionName",
                   files= [[path.parquet]] )

### Milvus Log

Attached are the logs of DataNode, which is in the CrashLoop.
[output.log](https://github.com/user-attachments/files/18281680/output.log)


### Anything else?

_No response_

[bigsheeper]

/assign

[bigsheeper]

[akmalmasud96]

The format of parquet file which i am importing is as follows,
key: str, embeddings: float16

and when we create the parquet using pymilvus LocalBulkWriter, the type of embeddings is uint8 for float16 vector, does milvus also expect uint8 type for parquet ?