Skip to content


Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time

Duplicacy Autobackup

build status latest version

Duplicacy Autobackup is a Docker image to easily perform automated backups. It uses duplicacy under the hood, and therefore supports:

  • Multiple storage backends: S3, Backblaze B2, Hubic, Dropbox, SFTP...
  • Client-side encryption
  • Deduplication
  • Multi-versioning
  • ... and more generally, all the features that duplicacy has.


The following environment variables can be used to configure the backup strategy.

  • BACKUP_NAME: The name of your backup (should be unique, e.g. prod-db-backups)
  • BACKUP_ENCRYPTION_KEY: An optional passphrase to encrypt your backups with before they are stored remotely.
  • BACKUP_SCHEDULE: Cron-like string to define the frequency at which backups should be made (e.g. 0 2 * * * for Every day at 2am). Note that this string should be indicated in the UTC timezone.
  • BACKUP_LOCATION: Duplicacy URI of where to store the backups.
    • S3: s3://
    • Backblaze B2: b2://my-bucket/
    • ...

Additionally, the directory you want to backup must be mounted to /data on the container.

You need to provide credentials for the storage provider your of your choice using the following environment variables:

  • Backblaze B2: B2_ID and B2_KEY
  • Dropbox: DROPBOX_TOKEN
  • Azure: AZURE_KEY
  • Google Cloud Datastore: GCD_TOKEN
  • Google Cloud Storage: GCS_TOKEN_FILE*

Environment variables marked with an asterix point to files. Those files must be mounted in the container so that they can be accessed from inside it.

If you want to execute an out of schedule backup, you can do so by running the script /app/ inside the container :

$ docker exec duplicacy-autobackup /app/ backup


Backup /var/lib/mysql to the S3 bucket xtof-db-backups in the AWS region eu-west-1 every night at 2:00am, and encrypt them with the passphrase correct horse battery staple:

$ docker run -d --name duplicacy-autobackup \
    -v /var/lib/mysql:/data \
    -e BACKUP_NAME='prod-db-backups' \
    -e BACKUP_LOCATION='s3://' \
    -e BACKUP_SCHEDULE='0 2 * * *' \
    -e BACKUP_ENCRYPTION_KEY='correct horse battery staple' \
    -e AWS_ACCESS_KEY_ID='AKIA...' \
    -e AWS_SECRET_KEY='...' \

Viewing and restoring backups

Backups are useless if you don't make sure they work. This shows the procedure to list files, versions, and restore a duplicacy backup made using duplicacy-autobackup.

  • Install Duplicacy: download the latest Duplicacy binary from its Github page, and put it in your path

  • cd to a directory where you'll restore your files, e.g. /tmp/restore

  • Run duplicacy init backup_name backup_location, where backup_name and backup_location correspond to the BACKUP_NAME and BACKUP_LOCATION environment variables of your setup.

    • If you used client-side encryption, add the -encrypt flag: duplicacy init -encrypt backup_name backup_location

    You will get a prompt asking for your storage provider's credentials, and, if applicable, your encryption key:

    Enter S3 Access Key ID: *****
    Enter S3 Secret Access Key: *************
    Enter storage password for s3://*******************
    The storage 's3://' has already been initialized
    Compression level: 100
    Average chunk size: 4194304
    Maximum chunk size: 16777216
    Minimum chunk size: 1048576
    Chunk seed: fc7e56fb91f8f66b01ba033ec6f7b128bcb3420c66a31468a4f3541407d569bd
    /tmp/restore will be backed up to s3:// with id db-backups
  • To list the versions of your backups, run:

    $ duplicacy list
    Storage set to s3://
    Enter storage password:*******************
    Snapshot db-backups revision 1 created at 2018-04-19 09:47 -hash
    Snapshot db-backups revision 2 created at 2018-04-19 09:48 
    Snapshot db-backups revision 3 created at 2018-04-19 09:49 
  • To view the files of a particular revision, run:

    $ duplicacy list -files -r 2  # 2 is the revision number
  • To restore in the current directory all the files matching *.txt of the revision 2 of the backup, run:

    $ duplicacy restore -r 2 '*.txt'
  • To restore in the current directory the whole revision 2 of your backup, run:

    $ duplicacy restore -ignore-owner -r 2

More: see Duplicacy's documentation.

Other options

You can have duplicacy-autobackup run a script before and after the backup process by mounting scripts on /scripts/ and /scripts/ For instance if you're backing up a MySQL database, this script can involve doing a mysqldump into /data/mydb.sql. If exits with a non-zero status code, the backup will not be performed until the next scheduled backup.

Use the following environment variables if you want to customize duplicacy's behavior.

  • BACKUP_IMMEDIATELY (yes/no): indicates if a backup should be performed immediately after the container is started. Equivalent to launching the container and then running docker exec duplicacy-autobackup /app/ backup. By default, no.
  • DUPLICACY_INIT_OPTIONS: options passed to duplicacy init the first time a backup is made. By default, -encrypt if BACKUP_ENCRYPTION_KEY is not empty.
  • DUPLICACY_BACKUP_OPTIONS: options passed to duplicacy backup when a backup is performed. By default: -threads 4 -stats. If you are backing up a hard drive (and not a SSD), it is recommended to use -threads 1 -stats instead (see here for more details).

Pruning old backups

Duplicacy offers an option to prune old backups. By default, duplicacy-autobackup does not perform any pruning. However, you can set the environment variables DUPLICACY_PRUNE_OPTIONS and PRUNE_SCHEDULE to perform automatic pruning. As an example, setting:

DUPLICACY_PRUNE_OPTIONS='-keep 0:360 -keep 30:180 -keep 7:30'
PRUNE_SCHEDULE='0 0 * * *'

Means that:

  • Every day at midnight, the pruning process runs
  • When the pruning process runs...
    • Any backup older than 1 year is deleted from the remote storage
    • Only 1 backup per 30 days is kept for backups between 180 days and 360 days old
    • Only 1 backup per 7 days is kept for backups between 7 days and 180 days old
    • 1 backup per day is kept for backups between 0 day and 7 days old

See the prune command details for further details.

Choosing the Duplicacy version

When building the container, you can choose the Duplicacy version that will be used in the container image. The build argument DUPLICACY_VERSION is available for that purpose, e.g.:

docker build --build-arg DUPLICACY_VERSION=2.1.0 -t christophetd/duplicacy-autobackup .


This project uses Duplicacy, which is free for personal use but requires purchasing a licence for non-trial commercial use. See the detailed terms here.


Feel free to open an issue for any suggestion or bug. You can also tweet @christophetd.