[Bug]: Standalone oomkilled after upgrading mater branch image

[ThreadDao]

Is there an existing issue for this?

• I have searched the existing issues

Environment

- Milvus version: e83ecd50 -> master-20240621-4e414fb7-amd64
- Deployment mode(standalone or cluster): standalone
- MQ type(rocksmq, pulsar or kafka):   rocksmq  
- SDK version(e.g. pymilvus v2.0.0rc2):
- OS(Ubuntu or CentOS): 
- CPU/Memory: 
- GPU: 
- Others:

Current Behavior

1. standalone pod zong-go-master-milvus-standalone-5cc75d965b-t46jr with resource config

    Limits:
      cpu:     8
      memory:  16Gi
    Requests:
      cpu:      5
      memory:   9Gi

2. upgrade image to master-20240621-4e414fb7-amd64 and pod oomkilled
3. upgrade standalone pod memory limit to 32Gi and pod is running

• pod memory usage

• metrics of zong-go-master

Expected Behavior

No response

Steps To Reproduce

No response

Milvus Log

pods:

zong-go-master-etcd-0                                             1/1     Running                           0              3d      10.104.25.21    4am-node30   <none>           <none>
zong-go-master-milvus-standalone-6cdbf578d-phtls                  1/1     Running                           0              10m     10.104.19.7     4am-node28   <none>           <none>
zong-go-master-minio-55956b966c-s48rt                             1/1     Running                           0              3d      10.104.34.72    4am-node37   <none>           <none>

Anything else?

No response

[yanliang567]

there are only 20 collections with totally 15k vectors loaded before upgrading, btw

[ThreadDao]

pyroscope of pod zong-go-master-milvus-standalone-6cdbf578d-phtls: http://10.100.36.158:4040/?query=zong-go-master-milvus-standalone.alloc_objects%7B%7D&from=1718955620&until=1718959425

[xiaofan-luan]

So we have 20collections with only 15K vectors each collection and 16GB standalone get OOM?

[sunby]

same issue: [Bug]: Upgrading from v2.3.15 to master-20240625-506a9152-amd64 due to DN and QN crash and one index