Perfect for data volumes like home partitions that are nearly full.
New mode: Files → Borg
- Creates LVM snapshot (consistent point-in-time state)
- Mounts snapshot to temporary location
- Borg backs up actual files (not empty blocks)
- Unmounts and cleans up snapshot
Benefits vs block-level backup:
- Space efficient: 305GB files vs 358GB entire LV
- Better compression: Borg works on real file data
- Superior deduplication: File-content based
- No snapshot space issues: Shorter operation time
- Easier file recovery: Browse/restore individual files
Use cases:
- Block-level: System volumes (root, boot) for exact state
- File-level: Data volumes (home) for space efficiency
GUI: Added 'Files → Borg' mode
CLI: Added 'files-to-borg' mode
Example:
sudo ./enhanced_simple_backup.sh files-to-borg /dev/internal-vg/home /repo --new-repo
This solves the 90% full home partition backup problem!
Problem: Fixed 500MB snapshot size was too small for large/active LVs,
causing I/O errors when snapshot space was exhausted during backup.
Solution: Calculate appropriate snapshot size based on LV size:
- Use 10% of LV size for snapshot space
- Minimum 1GB to handle reasonable activity
- Shows snapshot size in logs for transparency
Example:
- 20GB LV → 2GB snapshot
- 100GB LV → 10GB snapshot
- 500MB LV → 1GB snapshot (minimum)
This prevents snapshot overflow and resulting I/O errors during backup.
Changes:
- GUI: Calculate snapshot size before creating
- CLI: Same dynamic sizing logic
- Both show calculated snapshot size in output
- Handles both single LV and VG backup modes
Problem: VG→Borg was creating temp files for all LVs simultaneously,
causing 'no space left' errors when LVs are large.
Solution: Process LVs sequentially instead of simultaneously:
1. Create snapshot for one LV
2. Stream it directly to Borg (no temp files)
3. Remove snapshot immediately
4. Move to next LV
Changes:
- VG→Borg now creates separate archives per LV instead of one big archive
- Each LV gets its own archive: vg_internal-vg_lv_root_20241009_123456
- No temporary files needed - direct streaming
- Space-efficient: only one snapshot exists at a time
- More robust: failure of one LV doesn't affect others
Benefits:
- No more space issues regardless of LV sizes
- Faster cleanup between LVs
- Individual LV recovery possible
- Better error isolation
- Still preserves block-level backup benefits
BREAKING CHANGE: Borg backups now store raw block devices instead of files
Changes:
- LV→Borg: Stores snapshot as raw block device (.img file) in Borg
- VG→Borg: Stores each LV as separate .img files in Borg repository
- No more file-level mounting - preserves exact block-level state
- Uses dd | borg create --stdin-name for LV backups
- Creates temporary .img files for VG backups
- Maintains all filesystem metadata, boot sectors, etc.
- Better deduplication for similar block patterns
Benefits:
- Exact block-level restoration possible
- Preserves all filesystem metadata
- Better suited for system/boot volume backups
- Still gets Borg's compression, deduplication, encryption
- Clear difference between LV and VG modes
Now LV→Borg and VG→Borg have distinct, useful purposes:
- LV→Borg: Single logical volume as one block device
- VG→Borg: All logical volumes as separate block devices
- Add LV → Borg and VG → Borg backup modes
- GUI: Borg settings panel with repo path, encryption, passphrase
- CLI: Enhanced script with Borg options (--new-repo, --encryption, --passphrase)
- Automatic repository initialization for new repos
- Support for all Borg encryption modes (none, repokey, keyfile)
- Mount snapshots temporarily for file-level Borg backups
- Comprehensive cleanup of snapshots and mount points
- Updated documentation and examples
Benefits:
- Deduplication and compression
- Strong encryption support
- Incremental backups capability
- Space-efficient storage
- Still maintains simple snapshot → backup → cleanup workflow
- Move all old complex backup scripts to old_scripts/
- Archive previous documentation versions
- Clean up temporary files and debian packages
- Update README to focus on new simple system
- Keep only the enhanced simple backup system in main directory
Main directory now contains only:
- simple_backup_gui.py (GUI interface)
- enhanced_simple_backup.sh (CLI interface)
- list_drives.sh (helper)
- simple_backup.sh (basic CLI)
- SIMPLE_BACKUP_README.md (detailed docs)
- README.md (project overview)
- Add 3 backup modes: LV→LV, LV→Raw, VG→Raw
- Simple GUI with mode selection and dynamic target lists
- Command-line version with clear mode support
- Enhanced drive listing with mode-specific options
- Minimal logic: just snapshot → copy → cleanup
- No complex migration features that cause system issues
- Supports refreshing existing backups and creating fresh ones
- Fixed partition size calculation bugs in migrate_to_lvm.sh
- Added comprehensive error handling for USB drive stability
- Optimized data copy operations using cp -a for better performance
- Corrected mount point detection for encrypted home partitions
- Enhanced drive detection and exclusion logic
- Added proper size override mechanisms for manual intervention
- Improved filesystem creation and validation processes
- Complete toolset for external M.2 drive migration scenarios
Tested successfully on:
- Debian 13 Trixie Live USB environment
- 476GB external M.2 drives via USB 3.0
- Complex partition layouts (root/home/EFI + encryption)
- Large data transfers (314GB+ encrypted home directories)
- Added smart sync backup feature using rsync for incremental updates
- Implemented change analysis to recommend sync vs full clone
- Added GUI buttons for 'Smart Sync Backup' and 'Analyze Changes'
- Enhanced CLI with --sync and --analyze flags
- Smart sync provides 10-100x speed improvement for minor changes
- Maintains full system consistency while eliminating downtime
- Updated documentation with comprehensive smart sync guide
- All existing backup/restore functionality preserved
