A set of scripts for running Duplicacy on Unix systems
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.



This is a set of scripts that can be used to manage running Duplicacy(GitHub) on Unix systems. They were developed and tested under Linux but should work in any environment that complies with POSIX.

NOTE: Limitations in the current implementation of Duplicacy and the fact that Windows does not handle symbolic linking in a POSIX-like way makes these scripts unsuitable for that environment.


To install this package on your system:

Clone this repository: git clone https://github.com/markfeit/duplicacy-scripts.git

cd duplicacy-scripts

Download the Duplicacy binary from Gilbert Chen's release page or build one of your own and and place the file in duplicacy-scripts/duplicacy . Note that the license for Duplicacy imposes some restrictions on its use. Please abide by them.

Select a location where duplicacy-scripts, the Duplicacy configuration, its cache and log files will be kept. This location will be referred to as $DEST. The default is /opt/duplicacy.

Select the location which will form the root of the volume(s) to be backed up. This will be referred to as $ROOT. The default is /, which is suitable for most systems. (Specific parts of the filesystem may be included or excluded using Duplicacy's filter mechanism.)

Become root and execute:

  • make install to install using the defaults
  • make DEST=$DEST ROOT=$ROOT install to install using other directories.

Installation may be done as any other user, but be aware that this will limit the set of files backed up to those the user can read. In addition to installing these scripts, a .duplicacy file will be placed in $ROOT.

Set up Duplicacy by placing a preferences and optional filter file in $DEST/prefs Samples are provided in the prefs directory of the sources. These files are not installed by default.

Set up the scripts by editing $DEST/etc/settings.

At this point, backups and maintenance will be done automatically by cron.


Backups are run at 00:45 local time each morning. If there is another backup running (common when there is a long initial backup running), the newer backup will be aborted.

Logs of what happens during each backup and other matinenance activities are stored in $DEST/var/log. The latest logs or those for a specific date can be retrieved and read with $DEST/bin/logs.


Files can be restored by executing $DEST/bin/restore. Full documentation is (for now) in that file.


Daily (03:00):

  • Prune old snapshots according to the rules in etc/prune.
  • Remove log files older than 30 days.
  • Pull and update the software (if enabled in $DEST/etc/settings).

Weekly (Sunday at 03:15):

  • Fossilize and remove chunks that are no longer referenced.

  • Check integrity. This verifies that all chunks that should be present and attempts to resurrect missing chunks from fossils if possible. There is no attempt to download and verify the contents of the chunks. (Not implemented yet.)

Monthly (First Sunday at 03:30):

  • Nothing yet.