The difference between the dataset size of Redis resources and the size after export and restoration is too large

Hi Amrinder Singh,

I am glad to receive your reply, and I can confirm that this is also my doubt.

I tried to parse RDB files （about 40MB） using the redis check rdb command line tool. The size of the use-mem seen (about 200MB) is also inconsistent with the used_memory_dataset (about 1.6GB) information viewed through info in the running Redis resource.

Firstly, before cleaning up the number of keys in the early stage, I thought it might be a compression issue, but after cleaning up the big keys later, the remaining usage of this dataset made me very distressed.

The size of the RDB exported from Redis data structure through the check tool is inconsistent with the size in the info, and the size of the restored RDB is also inconsistent with the size seen in the check.

Will there be additional data structure optimization adjustments during the process of RDB reconstruction, such as the possibility of redundancy in the data structure filled in earlier stages, resulting in significant differences in optimization when exporting and refilling?

We are unable to view the complete configuration of Azure Redis, and I would like to know if there is a solution to resolve the current false occupation.