[Bug]: [benchmark]load memory usage more

[elstic]

Is there an existing issue for this?

• [X] I have searched the existing issues

Environment

- Milvus version:
- Deployment mode(standalone or cluster):standalone
- MQ type(rocksmq, pulsar or kafka):    
- SDK version(e.g. pymilvus v2.0.0rc2):
- OS(Ubuntu or CentOS): 
- CPU/Memory: 
- GPU: 
- Others:

Current Behavior

Save load memory pr After entering, load memory usage is higher instead . index : HNSW dataset : 6m 128d image :

Adjusting memory usage before load：6G ,image : 2.2.0-20230324-9c3c29db

Adjusting memory usage after load： 6.68G, image: xige-16-opt-mem-copy-aa2886b-20230403

Expected Behavior

Adjusted memory usage after load less than peak memory usage before load adjustment

Steps To Reproduce

No response

Milvus Log

No response

Anything else?

No response

[elstic]

[cluster] load memory usage more

Current Behavior

Adjusting memory usage before load：4.3G (querynode)

Adjusting memory usage after load： 5.5G (querynode)

[xiaofan-luan]

/assign @xige-16

[yanliang567]

/unassign

[binbinlv]

@xige-16 what is the progress now?

[xige-16]

2.2.0 has been verifyed , and can load more data

[elstic]

issue Validated, please be aware that: @yanliang567

After entering load pr, slightly more memory resources are needed when memory is sufficient, but more data can be loaded when resources are limited. See the table for details:

(standalone ， index : HNSW)

data_size\image	memory	v2.2.5	2.2.0-20230425-e820479b
6m	Sufficient resources	6.23G	6.81G
8m	2c9g	load failed	7.22G

[yanliang567]

@elstic so shall we close this issue now?

[elstic]

@elstic so shall we close this issue now?

Yes, I'll close this issue