When a duplicate stream is found, xtool removes the redundant data and replaces it with a small reference in a database file (e.g., xtool.bin ) .

# Find repeated chunks inside a single large file (e.g., database dump) xdelta3 -e --dedup -s /dev/null largefile largefile.dedup

-dd flag instructs xtool to scan for identical compressed streams (such as repeated assets in video games or media libraries). When xtool detects these duplicates, it removes the extra copies from the output file. During the restoration (decoding) process, only one instance of the stream is restored and then copied to all necessary locations, which avoids wasting CPU cycles on redundant decompression. encode.su Key Considerations Stream-Level Detection

Unlike fixed-block deduplication (e.g., 4KB blocks), xdelta3 uses (CDC) similar to LBFS or rsync.

However, for , it is a superior alternative, offering storage economics that standard dd cannot match.

# Decode (no dedup needed at decode time) xdelta3 -d -s image1.vmdk image2.delta image2_restored.vmdk