Add configurable S3 storage class for backups + archive restore workflow

What problem will this feature address?

Dokploy backups currently rely on destination defaults for S3 storage class, so users cannot choose storage tier per backup policy. This makes it hard to optimize cost and retention (for example, archive long-term backups while keeping short-term backups in faster tiers).

Describe the solution you'd like

Add optional, provider-aware S3 storage class support for both database backups and volume backups, and add an archive-restore workflow in restore screens.

Proposed behavior:

• Add optional storageClass field to backup and volume backup models.
• Show a storage class selector in backup forms with Use destination default option.
• Validate selected storage class server-side based on destination provider.
• Pass the selected class to rclone upload (--s3-storage-class) when valid.
• In restore views, show per-file:
  • storage class
  • restore availability (Ready, Restoring, Archived, Unknown)
  • readable-until time when available
• For archived objects, provide a “Request restore from AWS” action with:
  • retrieval tier (standard/priority/bulk)
  • restore lifetime (days)
  • clear status feedback after request

This should allow safer, cheaper storage strategies while making cold-storage restore flows understandable.

Describe alternatives you've considered

• Managing storage classes only at bucket/destination level: simpler but not flexible enough for mixed retention strategies.
• Manually running cloud-provider restore operations outside Dokploy: works as a workaround but is operationally heavy and error-prone.

Additional context

I already have an implementation/prototype branch that includes:

• schema/API/UI updates for storageClass
• provider-specific validation
• restore availability/status display
• archive restore request flow
• tests and migrations

I’d like maintainer feedback before finalizing/adjusting this approach, especially on:

• provider/class mapping expectations
• restore UX expectations for archived objects
• migration safety requirements for real-world upgrade paths

Will you send a PR to implement it?

Yes

Start by reviewing the prototype branch’s schema/API/UI updates, provider-specific validation, migrations, and tests, then trace the rclone upload path and restore views. Done means database and volume backups support optional provider-valid storage classes, while restore screens show archive status and support restore requests with feedback.