feat: Add file-level snapshot backup mode for space efficiency

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!
This commit is contained in:
root
2025-10-09 08:03:00 +02:00
parent f12d07be7f
commit 7708b98674
2 changed files with 292 additions and 28 deletions

View File

@@ -24,25 +24,28 @@ usage() {
echo " $0 vg-to-raw SOURCE_VG TARGET_DEVICE"
echo " $0 lv-to-borg SOURCE_LV REPO_PATH [--new-repo] [--encryption MODE] [--passphrase PASS]"
echo " $0 vg-to-borg SOURCE_VG REPO_PATH [--new-repo] [--encryption MODE] [--passphrase PASS]"
echo " $0 files-to-borg SOURCE_LV REPO_PATH [--new-repo] [--encryption MODE] [--passphrase PASS]"
echo ""
echo "Modes:"
echo " lv-to-lv - Update existing LV backup (SOURCE_LV → TARGET_LV)"
echo " lv-to-raw - Create fresh backup (SOURCE_LV → raw device)"
echo " vg-to-raw - Clone entire VG (SOURCE_VG → raw device)"
echo " lv-to-borg - Backup LV to Borg repository"
echo " vg-to-borg - Backup entire VG to Borg repository"
echo " lv-to-borg - Backup LV to Borg repository (block-level)"
echo " vg-to-borg - Backup entire VG to Borg repository (block-level)"
echo " files-to-borg - Backup LV files to Borg repository (file-level, space-efficient)"
echo ""
echo "Borg Options:"
echo " --new-repo Create new repository"
echo " --encryption MODE Encryption mode: none, repokey, keyfile (default: repokey)"
echo " --passphrase PASS Repository passphrase"
echo " --generous-snapshots Use 25% of LV size for snapshots (for very active systems)"
echo ""
echo "Examples:"
echo " $0 lv-to-lv /dev/internal-vg/root /dev/backup-vg/root"
echo " $0 lv-to-raw /dev/internal-vg/root /dev/sdb"
echo " $0 vg-to-raw internal-vg /dev/sdb"
echo " $0 lv-to-borg /dev/internal-vg/root /path/to/borg/repo --new-repo"
echo " $0 vg-to-borg internal-vg /path/to/borg/repo --encryption repokey --passphrase mypass"
echo " $0 files-to-borg /dev/internal-vg/home /path/to/borg/repo --new-repo"
echo ""
echo "List available sources/targets:"
echo " ./list_drives.sh"
@@ -95,6 +98,7 @@ shift 3
NEW_REPO=false
ENCRYPTION="repokey"
PASSPHRASE=""
GENEROUS_SNAPSHOTS=false
while [ $# -gt 0 ]; do
case "$1" in
@@ -110,6 +114,10 @@ while [ $# -gt 0 ]; do
PASSPHRASE="$2"
shift 2
;;
--generous-snapshots)
GENEROUS_SNAPSHOTS=true
shift
;;
*)
error "Unknown option: $1"
;;
@@ -123,7 +131,7 @@ fi
# Validate mode
case "$MODE" in
"lv-to-lv"|"lv-to-raw"|"vg-to-raw"|"lv-to-borg"|"vg-to-borg")
"lv-to-lv"|"lv-to-raw"|"vg-to-raw"|"lv-to-borg"|"vg-to-borg"|"files-to-borg")
;;
*)
error "Invalid mode: $MODE"
@@ -300,10 +308,22 @@ case "$MODE" in
# Get LV size to determine appropriate snapshot size
LV_SIZE_BYTES=$(lvs --noheadings -o lv_size --units b "$SOURCE" | tr -d ' ' | sed 's/B$//')
# Use 10% of LV size or minimum 1GB for snapshot
SNAPSHOT_SIZE_BYTES=$((LV_SIZE_BYTES / 10))
if [ $SNAPSHOT_SIZE_BYTES -lt 1073741824 ]; then
SNAPSHOT_SIZE_BYTES=1073741824 # 1GB minimum
# Use different percentages based on options
if [ "$GENEROUS_SNAPSHOTS" = true ]; then
SNAPSHOT_SIZE_BYTES=$((LV_SIZE_BYTES / 4)) # 25%
MIN_SIZE=$((2 * 1024 * 1024 * 1024)) # 2GB minimum
else
# Auto mode: 10% normally, 15% for large LVs
if [ $LV_SIZE_BYTES -gt $((50 * 1024 * 1024 * 1024)) ]; then
SNAPSHOT_SIZE_BYTES=$((LV_SIZE_BYTES * 15 / 100)) # 15% for >50GB
else
SNAPSHOT_SIZE_BYTES=$((LV_SIZE_BYTES / 10)) # 10% normally
fi
MIN_SIZE=$((1024 * 1024 * 1024)) # 1GB minimum
fi
if [ $SNAPSHOT_SIZE_BYTES -lt $MIN_SIZE ]; then
SNAPSHOT_SIZE_BYTES=$MIN_SIZE
fi
SNAPSHOT_SIZE_GB=$((SNAPSHOT_SIZE_BYTES / 1073741824))
SNAPSHOT_SIZE="${SNAPSHOT_SIZE_GB}G"
@@ -411,10 +431,22 @@ case "$MODE" in
# Get LV size to determine appropriate snapshot size
LV_SIZE_BYTES=$(lvs --noheadings -o lv_size --units b "$LV_PATH" | tr -d ' ' | sed 's/B$//')
# Use 10% of LV size or minimum 1GB for snapshot
SNAPSHOT_SIZE_BYTES=$((LV_SIZE_BYTES / 10))
if [ $SNAPSHOT_SIZE_BYTES -lt 1073741824 ]; then
SNAPSHOT_SIZE_BYTES=1073741824 # 1GB minimum
# Use different percentages based on options
if [ "$GENEROUS_SNAPSHOTS" = true ]; then
SNAPSHOT_SIZE_BYTES=$((LV_SIZE_BYTES / 4)) # 25%
MIN_SIZE=$((2 * 1024 * 1024 * 1024)) # 2GB minimum
else
# Auto mode: 10% normally, 15% for large LVs
if [ $LV_SIZE_BYTES -gt $((50 * 1024 * 1024 * 1024)) ]; then
SNAPSHOT_SIZE_BYTES=$((LV_SIZE_BYTES * 15 / 100)) # 15% for >50GB
else
SNAPSHOT_SIZE_BYTES=$((LV_SIZE_BYTES / 10)) # 10% normally
fi
MIN_SIZE=$((1024 * 1024 * 1024)) # 1GB minimum
fi
if [ $SNAPSHOT_SIZE_BYTES -lt $MIN_SIZE ]; then
SNAPSHOT_SIZE_BYTES=$MIN_SIZE
fi
SNAPSHOT_SIZE_GB=$((SNAPSHOT_SIZE_BYTES / 1073741824))
SNAPSHOT_SIZE="${SNAPSHOT_SIZE_GB}G"
@@ -445,6 +477,101 @@ case "$MODE" in
log "VG to Borg backup completed successfully"
;;
"files-to-borg")
# Files to Borg repository backup
if [ ! -e "$SOURCE" ]; then
error "Source LV does not exist: $SOURCE"
fi
# Extract VG and LV names
VG_NAME=$(lvs --noheadings -o vg_name "$SOURCE" | tr -d ' ')
LV_NAME=$(lvs --noheadings -o lv_name "$SOURCE" | tr -d ' ')
SNAPSHOT_NAME="${LV_NAME}_files_snap"
SNAPSHOT_PATH="/dev/$VG_NAME/$SNAPSHOT_NAME"
# Get LV size to determine appropriate snapshot size
LV_SIZE_BYTES=$(lvs --noheadings -o lv_size --units b "$SOURCE" | tr -d ' ' | sed 's/B$//')
# Use different percentages based on options
if [ "$GENEROUS_SNAPSHOTS" = true ]; then
SNAPSHOT_SIZE_BYTES=$((LV_SIZE_BYTES / 4)) # 25%
MIN_SIZE=$((2 * 1024 * 1024 * 1024)) # 2GB minimum
else
# Auto mode: 10% normally, 15% for large LVs
if [ $LV_SIZE_BYTES -gt $((50 * 1024 * 1024 * 1024)) ]; then
SNAPSHOT_SIZE_BYTES=$((LV_SIZE_BYTES * 15 / 100)) # 15% for >50GB
else
SNAPSHOT_SIZE_BYTES=$((LV_SIZE_BYTES / 10)) # 10% normally
fi
MIN_SIZE=$((1024 * 1024 * 1024)) # 1GB minimum
fi
if [ $SNAPSHOT_SIZE_BYTES -lt $MIN_SIZE ]; then
SNAPSHOT_SIZE_BYTES=$MIN_SIZE
fi
SNAPSHOT_SIZE_GB=$((SNAPSHOT_SIZE_BYTES / 1073741824))
SNAPSHOT_SIZE="${SNAPSHOT_SIZE_GB}G"
info "This will backup LV files to Borg repository (space-efficient)"
echo ""
echo -e "${YELLOW}Source LV: $SOURCE${NC}"
echo -e "${YELLOW}Repository: $TARGET${NC}"
echo -e "${BLUE}Snapshot size: $SNAPSHOT_SIZE${NC}"
echo -e "${BLUE}Mode: File-level backup (skips empty blocks)${NC}"
if [ "$NEW_REPO" = true ]; then
echo -e "${BLUE}Will create new repository${NC}"
else
echo -e "${BLUE}Will add to existing repository${NC}"
fi
echo -e "${BLUE}Encryption: $ENCRYPTION${NC}"
echo ""
read -p "Continue? (yes/no): " confirm
if [ "$confirm" != "yes" ]; then
echo "Backup cancelled."
exit 0
fi
# Initialize repository if needed
if [ "$NEW_REPO" = true ]; then
log "Creating new Borg repository: $TARGET"
if [ "$ENCRYPTION" = "none" ]; then
borg init --encryption=none "$TARGET" || error "Failed to initialize repository"
else
borg init --encryption="$ENCRYPTION" "$TARGET" || error "Failed to initialize repository"
fi
log "Repository initialized successfully"
fi
log "Creating snapshot of source LV"
lvcreate -L"$SNAPSHOT_SIZE" -s -n "$SNAPSHOT_NAME" "$SOURCE" || error "Failed to create snapshot"
# Create temporary mount point
TEMP_MOUNT=$(mktemp -d -t borg_files_backup_XXXXXX)
# Mount snapshot
log "Mounting snapshot to $TEMP_MOUNT"
mount "$SNAPSHOT_PATH" "$TEMP_MOUNT" || error "Failed to mount snapshot"
# Create Borg archive
ARCHIVE_NAME="files_${LV_NAME}_$(date +%Y%m%d_%H%M%S)"
log "Creating Borg archive (file-level): $ARCHIVE_NAME"
log "Backing up files from mounted snapshot..."
log "This is space-efficient and skips empty blocks"
borg create --progress --stats --compression auto,zstd "$TARGET::$ARCHIVE_NAME" "$TEMP_MOUNT" || error "Borg file-level backup failed"
log "File-level Borg backup completed successfully"
# Cleanup
log "Cleaning up mount point and snapshot"
umount "$TEMP_MOUNT" || warn "Failed to unmount"
rmdir "$TEMP_MOUNT"
lvremove -f "$SNAPSHOT_PATH" || warn "Failed to remove snapshot"
SNAPSHOT_PATH=""
log "Files to Borg backup completed successfully"
;;
esac
echo ""

View File

@@ -45,12 +45,15 @@ class SimpleBackupGUI:
ttk.Radiobutton(mode_frame, text="Entire VG → Device",
variable=self.mode_var, value="vg_to_raw",
command=self.on_mode_change).pack(side=tk.LEFT, padx=5)
ttk.Radiobutton(mode_frame, text="LV → Borg Repo",
ttk.Radiobutton(mode_frame, text="LV → Borg (block)",
variable=self.mode_var, value="lv_to_borg",
command=self.on_mode_change).pack(side=tk.LEFT, padx=5)
ttk.Radiobutton(mode_frame, text="VG → Borg Repo",
ttk.Radiobutton(mode_frame, text="VG → Borg (block)",
variable=self.mode_var, value="vg_to_borg",
command=self.on_mode_change).pack(side=tk.LEFT, padx=5)
ttk.Radiobutton(mode_frame, text="Files → Borg",
variable=self.mode_var, value="files_to_borg",
command=self.on_mode_change).pack(side=tk.LEFT, padx=5)
# Source selection
ttk.Label(main_frame, text="Source:").grid(row=1, column=0, sticky=tk.W, pady=5)
@@ -88,14 +91,32 @@ class SimpleBackupGUI:
values=["none", "repokey", "keyfile"], state="readonly", width=15)
encryption_combo.grid(row=2, column=1, sticky=tk.W, pady=2)
# Snapshot size option
ttk.Label(self.borg_frame, text="Snapshot Size:").grid(row=3, column=0, sticky=tk.W, pady=2)
self.snapshot_size_var = tk.StringVar(value="auto")
snapshot_combo = ttk.Combobox(self.borg_frame, textvariable=self.snapshot_size_var,
values=["auto", "conservative", "generous"], state="readonly", width=15)
snapshot_combo.grid(row=3, column=1, sticky=tk.W, pady=2)
# Help text for snapshot sizes
help_label = ttk.Label(self.borg_frame, text="auto: smart sizing, conservative: 5%, generous: 25%",
font=("TkDefaultFont", 8), foreground="gray")
help_label.grid(row=3, column=2, sticky=tk.W, padx=(5, 0), pady=2)
# Read-only mode for full filesystems
self.readonly_mode = tk.BooleanVar()
ttk.Checkbutton(self.borg_frame, text="Read-only mode (for very full filesystems)",
variable=self.readonly_mode).grid(row=5, column=0, columnspan=3, sticky=tk.W, pady=2)
# Passphrase
ttk.Label(self.borg_frame, text="Passphrase:").grid(row=3, column=0, sticky=tk.W, pady=2)
ttk.Label(self.borg_frame, text="Passphrase:").grid(row=4, column=0, sticky=tk.W, pady=2)
self.passphrase_var = tk.StringVar()
passphrase_entry = ttk.Entry(self.borg_frame, textvariable=self.passphrase_var, show="*", width=40)
passphrase_entry.grid(row=3, column=1, sticky=(tk.W, tk.E), pady=2)
passphrase_entry.grid(row=4, column=1, sticky=(tk.W, tk.E), pady=2)
# Configure borg frame grid
self.borg_frame.columnconfigure(1, weight=1)
self.borg_frame.columnconfigure(2, weight=0)
# Refresh button
ttk.Button(main_frame, text="Refresh", command=self.refresh_drives).grid(row=4, column=0, pady=10)
@@ -165,6 +186,26 @@ class SimpleBackupGUI:
self.log_text.see(tk.END)
self.root.update_idletasks()
def calculate_snapshot_size(self, lv_size_bytes):
"""Calculate appropriate snapshot size based on LV size and user preference"""
mode = self.snapshot_size_var.get()
if mode == "conservative":
# 5% of LV size, minimum 1GB
snapshot_size_bytes = max(lv_size_bytes // 20, 1024**3)
elif mode == "generous":
# 25% of LV size, minimum 2GB
snapshot_size_bytes = max(lv_size_bytes // 4, 2 * 1024**3)
else: # auto
# 10% of LV size, minimum 1GB, but increase for very large LVs
base_percentage = 10
if lv_size_bytes > 50 * 1024**3: # >50GB
base_percentage = 15 # Use 15% for large, potentially active LVs
snapshot_size_bytes = max(lv_size_bytes * base_percentage // 100, 1024**3)
snapshot_size_gb = snapshot_size_bytes // (1024**3)
return f"{snapshot_size_gb}G"
def run_command(self, cmd, show_output=True):
"""Run a command and return result"""
try:
@@ -254,7 +295,7 @@ class SimpleBackupGUI:
else:
self.log("No target LVs found")
self.target_combo['values'] = []
elif mode in ["lv_to_borg", "vg_to_borg"]:
elif mode in ["lv_to_borg", "vg_to_borg", "files_to_borg"]:
# No target combo for Borg modes - repository path is used
self.target_combo['values'] = []
self.log("Using Borg repository path instead of target device")
@@ -283,7 +324,7 @@ class SimpleBackupGUI:
mode = self.mode_var.get()
# Validate inputs based on mode
if mode in ["lv_to_borg", "vg_to_borg"]:
if mode in ["lv_to_borg", "vg_to_borg", "files_to_borg"]:
if not self.source_var.get() or not self.repo_path_var.get():
messagebox.showerror("Error", "Please select source and repository path")
return
@@ -299,14 +340,22 @@ class SimpleBackupGUI:
# Build confirmation message for Borg
if mode == "lv_to_borg":
msg = f"Borg backup of LV:\n\nSource LV: {source}\nRepository: {repo_path}\n\n"
msg = f"Borg backup of LV (block-level):\n\nSource LV: {source}\nRepository: {repo_path}\n\n"
if self.create_new_repo.get():
msg += "This will create a new Borg repository.\n"
else:
msg += "This will add to existing Borg repository.\n"
msg += f"Encryption: {self.encryption_var.get()}\n\nContinue?"
elif mode == "files_to_borg":
msg = f"Borg backup of LV (file-level):\n\nSource LV: {source}\nRepository: {repo_path}\n\n"
msg += "This will backup files from the LV (space-efficient).\n"
if self.create_new_repo.get():
msg += "This will create a new Borg repository.\n"
else:
msg += "This will add to existing Borg repository.\n"
msg += f"Encryption: {self.encryption_var.get()}\n\nContinue?"
else: # vg_to_borg
msg = f"Borg backup of entire VG:\n\nSource VG: {source}\nRepository: {repo_path}\n\n"
msg = f"Borg backup of entire VG (block-level):\n\nSource VG: {source}\nRepository: {repo_path}\n\n"
if self.create_new_repo.get():
msg += "This will create a new Borg repository.\n"
else:
@@ -362,6 +411,8 @@ class SimpleBackupGUI:
self.backup_vg_to_raw(source, target)
elif mode == "lv_to_borg":
self.backup_lv_to_borg(source, target)
elif mode == "files_to_borg":
self.backup_files_to_borg(source, target)
elif mode == "vg_to_borg":
self.backup_vg_to_borg(source, target)
@@ -522,10 +573,7 @@ class SimpleBackupGUI:
success, lv_size_output = self.run_command(f"lvs --noheadings -o lv_size --units b {source_lv}", show_output=False)
if success:
lv_size_bytes = int(lv_size_output.strip().replace('B', ''))
# Use 10% of LV size or minimum 1GB for snapshot
snapshot_size_bytes = max(lv_size_bytes // 10, 1024**3) # Min 1GB
snapshot_size_gb = snapshot_size_bytes // (1024**3)
snapshot_size = f"{snapshot_size_gb}G"
snapshot_size = self.calculate_snapshot_size(lv_size_bytes)
else:
snapshot_size = "2G" # Default fallback
@@ -624,10 +672,7 @@ class SimpleBackupGUI:
success, lv_size_output = self.run_command(f"lvs --noheadings -o lv_size --units b {lv_path}", show_output=False)
if success:
lv_size_bytes = int(lv_size_output.strip().replace('B', ''))
# Use 10% of LV size or minimum 1GB for snapshot
snapshot_size_bytes = max(lv_size_bytes // 10, 1024**3) # Min 1GB
snapshot_size_gb = snapshot_size_bytes // (1024**3)
snapshot_size = f"{snapshot_size_gb}G"
snapshot_size = self.calculate_snapshot_size(lv_size_bytes)
else:
snapshot_size = "2G" # Default fallback
@@ -692,6 +737,98 @@ class SimpleBackupGUI:
self.current_snapshot = None
def backup_files_to_borg(self, source_lv, repo_path):
"""Backup LV files to Borg repository (file-level via snapshot mount)"""
self.log("Mode: LV to Borg Repository (file-level backup)")
self.log("This will mount an LV snapshot and backup files (space-efficient)")
# Set up environment for Borg
borg_env = os.environ.copy()
if self.passphrase_var.get():
borg_env['BORG_PASSPHRASE'] = self.passphrase_var.get()
# Initialize repository if needed
if self.create_new_repo.get():
self.log(f"Creating new Borg repository: {repo_path}")
encryption = self.encryption_var.get()
if encryption == "none":
init_cmd = f"borg init --encryption=none {repo_path}"
else:
init_cmd = f"borg init --encryption={encryption} {repo_path}"
result = subprocess.run(init_cmd, shell=True, capture_output=True, text=True, env=borg_env)
if result.returncode != 0:
raise Exception(f"Failed to initialize Borg repository: {result.stderr}")
self.log("Repository initialized successfully")
# Create snapshot
vg_name = source_lv.split('/')[2]
lv_name = source_lv.split('/')[3]
snapshot_name = f"{lv_name}_files_snap"
self.current_snapshot = f"/dev/{vg_name}/{snapshot_name}"
# Get LV size to determine appropriate snapshot size
success, lv_size_output = self.run_command(f"lvs --noheadings -o lv_size --units b {source_lv}", show_output=False)
if success:
lv_size_bytes = int(lv_size_output.strip().replace('B', ''))
snapshot_size = self.calculate_snapshot_size(lv_size_bytes)
else:
snapshot_size = "2G" # Default fallback
self.log(f"Creating snapshot: {snapshot_name} (size: {snapshot_size})")
success, output = self.run_command(f"lvcreate -L{snapshot_size} -s -n {snapshot_name} {source_lv}")
if not success:
raise Exception(f"Failed to create snapshot: {output}")
# Create temporary mount point
import tempfile
temp_mount = tempfile.mkdtemp(prefix="borg_files_backup_")
try:
# Mount the snapshot
self.log(f"Mounting snapshot to {temp_mount}")
success, output = self.run_command(f"mount {self.current_snapshot} {temp_mount}")
if not success:
raise Exception(f"Failed to mount snapshot: {output}")
# Create Borg backup of files
archive_name = f"files_{lv_name}_{time.strftime('%Y%m%d_%H%M%S')}"
self.log(f"Creating Borg archive (file-level): {archive_name}")
self.log("Backing up files from mounted snapshot...")
self.log("This is space-efficient and skips empty blocks")
borg_cmd = f"borg create --progress --stats --compression auto,zstd {repo_path}::{archive_name} {temp_mount}"
# Run borg with environment
process = subprocess.Popen(borg_cmd, shell=True, stdout=subprocess.PIPE,
stderr=subprocess.STDOUT, text=True, env=borg_env)
# Stream output
for line in process.stdout:
self.log(line.strip())
process.wait()
if process.returncode != 0:
raise Exception("Borg file-level backup failed")
self.log("File-level Borg backup completed successfully")
finally:
# Cleanup: unmount and remove temp directory
self.log("Cleaning up mount point")
self.run_command(f"umount {temp_mount}", show_output=False)
os.rmdir(temp_mount)
# Remove snapshot
self.log("Cleaning up snapshot")
success, output = self.run_command(f"lvremove -f {self.current_snapshot}")
if not success:
self.log(f"Warning: Failed to remove snapshot: {output}")
else:
self.log("Snapshot cleaned up")
self.current_snapshot = None
def cleanup_on_error(self):
"""Clean up on error"""
if self.current_snapshot: