br node OOM when running restore cmd manullay inside a pod with 56GB memory

## Bug Report

Please answer these questions before submitting your issue. Thanks!

### 1. Minimal reproduce step (Required)
1. deploy a br pod with below configuration in the GCP n2-standard-16 host.
```
requests:
cpu: 15000m
memory: 56Gi
limits:
cpu: 16000m
memory: 64Gi
```
2. connect to the pod via shell, and run br restore command to restore a br backup (no stats in the backup) of 300TB cluster to a new TiDB TiDB cluster with 100 TiKV node

### 2. What did you expect to see? (Required)
br node should not OOM.

### 3. What did you see instead (Required)
br node OOM.
When repeat the testing with 128gb br node, br node doesn't OOM.

### 4. What is your TiDB version? (Required)
v7.6.0

Start with the BR restore command in the reported TiDB v7.6.0 environment and reproduce it in a pod with 56Gi memory while restoring the 300TB backup to a 100-TiKV cluster. Investigate the cause of the BR node OOM and define the resource behavior expected under that configuration; done means the restore no longer OOMs with 56Gi memory.

Written by the indexing model from the issue text.