br restore not proceeded from checkpoint and user is prompted to drop system tables

[fubinzh]

Bug Report

Please answer these questions before submitting your issue. Thanks!

1. Minimal reproduce step (Required)

1. First br restore failed (exit code 137, might be due to br OOM.)
2. Run the br restore command again

2. What did you expect to see? (Required)

br restore should be proceeded from checkpoint.

3. What did you see instead (Required)

br restore failed due to "cluster is not fresh", user is prompted to drop databases/tables and many system tables are included which could be dangerous to drop.

4. What is your TiDB version? (Required)

[root@br-0 tmp]# /br -V                                   
Release Version: v7.6.0                                   
Git Commit Hash: 6701bf1527c7386a82396a1ecfb5322872daa2b3 
Git Branch: heads/refs/tags/v7.6.0                        
Go Version: go1.21.5                                      
UTC Build Time: 2024-01-08 12:55:13                       
Race Enabled: false        

[fubinzh]

/component br
/severity major