dd if=/dev/sdX1 of=mbr_backup.img bs=1M count=10 mkfs.exfat /dev/sdX1 dd if=mbr_backup.img of=/dev/sdX1 bs=1M count=10 conv=notrunc # This preserves cache if it starts after 10MB # Use mkntfs with --preserve (specific to ntfs-3g tools) mkntfs -Q -F /dev/sdX1 --preserve # The -Q (quick) and -F (force) skip bad block checks; --preserve keeps existing data clusters. Step 5: Verify Cache Integrity After Preparation After the "hold" operation, the drive should be ready—new file system, old cache intact. Verify:
echo "Step 4: Restoring header and unlocking cache..." dd if=$TEMP_BACKUP of=$DEVICE bs=1M count=20 conv=notrunc mount $DEVICE /mnt/new_drive
echo "Step 3: Recreating file system (exFAT or NTFS)..." read -p "Format as exFAT or NTFS? " FS if [ "$FS" == "exFAT" ]; then mkfs.exfat $DEVICE -n CACHE_DRIVE -v else mkfs.ntfs -Q -F $DEVICE --preserve -n CACHE_DRIVE fi
# Shrink NTFS from the end (keeps cache safe at the start) ntfsresize -s 120G /dev/sdX1 --no-action # Then adjust partition table with fdisk Most mkfs commands destroy data. However, you can use a hold pattern: For exFAT: # Create new exFAT but skip zeroing the cache clusters mkfs.exfat /dev/sdX1 -n MYDRIVE -v --keep-existing-files # (Note: --keep-existing-files is not standard in all mkfs.exfat; use dd workaround instead) Alternative dd workaround – backup first 10MB of drive (where FS lives), format, restore cache: