Safety and reliability: timestamped snapshot names, safer LUKS open via stdin, read-only filesystem-aware mounts, VG free-space preflight in GUI; CLI strict mode and file-level snapshot sizing; README updates
This commit is contained in:
@@ -94,6 +94,10 @@ sudo ./enhanced_simple_backup.sh vg-to-borg internal-vg /path/to/borg/repo --pas
|
||||
- Automatic snapshot cleanup on errors
|
||||
- Emergency stop functionality (GUI)
|
||||
- Input validation and error handling
|
||||
- Timestamped snapshot names to avoid collisions on retries
|
||||
- Read-only snapshot mounts with filesystem-aware safety flags (ext4: noload, xfs: norecovery)
|
||||
- Strict shell mode in CLI (set -euo pipefail) for reliable error propagation
|
||||
- Pre-flight checks for VG free space before creating snapshots
|
||||
|
||||
## What This System Does NOT Do
|
||||
|
||||
@@ -115,6 +119,11 @@ Previous versions of this project included complex migration logic that occasion
|
||||
✅ Predictable behavior
|
||||
✅ No hidden "smart" features
|
||||
|
||||
Notes and limitations:
|
||||
- Multi-PV VGs: Raw clone and VG→Borg flows require careful handling. If a VG spans multiple PVs, the simple raw clone path is not supported; use block-level per-LV backups or full-disk cloning tools.
|
||||
- File-level backups are for data safety, not bootability. Use block-level backups for exact bootable images.
|
||||
- Ensure the Borg repository is not located on the same LV being snapshotted.
|
||||
|
||||
## Recovery
|
||||
|
||||
### From LV Backup
|
||||
|
||||
Reference in New Issue
Block a user