@LoveYou3000, thanks for your PR, but I'm curious why you wanna change the num_shards for Milvus. Given that each shard in Milvus can handle at least 200 million data rows and that the public datasets provided by vdbbench are significantly smaller, we did not include an option for num_shards when designing the Milvus client.

Are you testing with a much larger dataset, or do you have a specific need for extremely high write speeds? Otherwise, we generally do not recommend increasing the number of shards.

We are doing a search performance test for the milvus cluster and want to compare the impact of different num_shards on the retrieval QPS.
BTW, my dataset is openai-large-10M

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: alwayslove2013, LoveYou3000
To complete the pull request process, please assign xuanyang-cn after the PR has been reviewed.
You can assign the PR to them by writing /assign @xuanyang-cn in a comment when ready.

The full list of commands accepted by this bot can be found here.

@LoveYou3000 Thanks for your willingness to test Milvus more comprehensively! We would love to hear your testing results, as they could be very valuable to the community.

Just a gentle reminder: considering the size of your test dataset, a single shard should be sufficient to handle it effectively.