In aws opensearch, document delete is keep occuring 24hours with same number

Versions (relevant - OpenSearch/Dashboard/Server OS/Browser):
opensearch domain seoul region 2.11

Describe the issue:
I create 2 data node in one cluster. (Not production, its dev)

I only use the cluster for searching docs and using time is 9:00am - 6pm.

However, after 6pm, all 24hours, when i check the matrics in console,

Documentdelete is occured continously with same number of docs, and memeory usage is keep almost 90%, cpu usage also keep almost 20%, and query per min also 2~3 every 15min, and jvm young garbage collect count is keep increasing every 15min. But 0 count old garbage collect count.

Im worry about why the above situation happened even in the night time i dont use

Configuration:

Relevant Logs or Screenshots: