[IMPROVEMENT] Backup and restore data Integrity check and repair

[derekbit]

Is your improvement request related to a feature? Please describe

During backing up the volume, the checksum of each data block is calculated. The checksums of the blocks are also verified during restore. To improve the data integrity more, we can consider store and verify the checksum of a volume or a snapshot during backup and restore. This will help reconfirm the restored volume is correct.

Additionally, each block is only backed up once by Longhorn because of the nature of the deduplication. Any intentional or unintentional corruption in backup target will lead to the restore failure. To address the issue, we probably can introduce some automatic repair mechanisms such as ECC to help fix the corruption.

Describe the solution you'd like

A clear and concise description of what you want to happen.

Describe alternatives you've considered

A clear and concise description of any alternative solutions or features you've considered.

Additional context

Add any other context or screenshots about the feature request here.