Backing up and moving Weblate¶
Automated backup using BorgBackup¶
New in version 3.9.
Weblate has built-in support for creating service backups using BorgBackup. Borg creates space-effective encrypted backups which can be safely stored in the cloud. The backups can be controlled in the management interface from the Backups tab.
Changed in version 4.4.1: Both PostgreSQL and MySQL/MariaDB databases are included in the automated backups.
The backups using Borg are incremental and Weblate is configured to keep following backups:
Daily backups for 14 days back
Weekly backups for 8 weeks back
Monthly backups for 6 months back
Borg encryption key¶
BorgBackup creates encrypted backups and you wouldn’t be able to restore them without the passphrase. The passphrase is generated when adding a new backup service and you should copy it and keep it in a secure place.
If you are using Weblate provisioned backup storage, please backup your private SSH key too, as it’s used to access your backups.
Weblate provisioned backup storage¶
The easiest way of backing up your Weblate instance is purchasing the backup service at weblate.org. This is how you get it running:
Purchase the Backup service on https://weblate.org/support/#backup.
Enter the obtained key in the management interface, see Integrating support.
Weblate connects to the cloud service and obtains access info for the backups.
Turn on the new backup configuration from the Backups tab.
Backup your Borg credentials to be able to restore the backups, see Borg encryption key.
The manual step of turning everything on is there for your safety. Without your consent no data is sent to the backup repository obtained through the registration process.
Using custom backup storage¶
You can also use your own storage for the backups. SSH can be used to store backups in the remote destination, the target server needs to have BorgBackup installed.
General in the Borg documentation
It is recommended to specify the absolute path for the local backup, for example /path/to/backup. The directory has to be writable by the user running Weblate (see Filesystem permissions). If it doesn’t exist, Weblate attempts to create it but needs the appropriate permissions to do so.
When running Weblate in Docker, please ensure the backup location is exposed as a volume from the Weblate container. Otherwise the backups will be discarded by Docker upon restarting the container it is in.
One option is to place backups into an existing volume, for example
/app/data/borgbackup. This is an existing volume in the container.
You can also add a new container for the backups in the Docker Compose file
for example by using
services: weblate: volumes: - /home/weblate/data:/app/data - /home/weblate/borgbackup:/borgbackup
The directory where backups will be stored have to be owned by UID 1000, otherwise Weblate won’t be able to write the backups there.
For creating remote backups, you will have to install BorgBackup onto another server that’s accessible for your Weblate deployment via SSH using the Weblate SSH key:
Prepare a server where your backups will be stored.
Install the SSH server on it (you will get it by default with most Linux distributions).
Choose an existing user or create a new user that will be used for backing up.
Add Weblate SSH key to the user so that Weblate can SSH to the server without a password (see Weblate SSH key).
Configure the backup location in Weblate as
Weblate provisioned backup storage provides you automated remote backups without any effort.
Restoring from BorgBackup¶
Restore access to your backup repository and prepare your backup passphrase.
List all the backups on the server using
borg list REPOSITORY.
Restore the desired backup to the current directory using
borg extract REPOSITORY::ARCHIVE.
Restore the database from the SQL dump placed in the
backupdirectory in the Weblate data dir (see Dumped data for backups).
When using Docker container, the settings file is already included in the container and you should restore the original environment variables. The
environment.ymlfile might help you with this (see Dumped data for backups).
Copy the whole restored data dir to the location configured by
When using Docker container place the data into the data volume, see Docker container volumes.
Please make sure the files have correct ownership and permissions, see Filesystem permissions.
The Borg session might look like this:
$ borg list /tmp/xxx Enter passphrase for key /tmp/xxx: 2019-09-26T14:56:08 Thu, 2019-09-26 14:56:08 [de0e0f13643635d5090e9896bdaceb92a023050749ad3f3350e788f1a65576a5] $ borg extract /tmp/xxx::2019-09-26T14:56:08 Enter passphrase for key /tmp/xxx:
Depending on what you want to save, back up the type of data Weblate stores in each respective place.
If you are doing the manual backups, you might want to
silence Weblate’s warning about a lack of backups by adding
WEBLATE_SILENCED_SYSTEM_CHECKS for Docker.
The actual storage location depends on your database setup.
The database is the most important storage. Set up regular backups of your database. Without the database, all the translations are gone.
Native database backup¶
The recommended approach is to save a dump of the database using database-native tools such as pg_dump or mysqldump. It usually performs better than Django backup, and it restores complete tables with all their data.
You can restore this backup in a newer Weblate release, it will perform all the
necessary migrations when running in
migrate. Please consult
Upgrading Weblate on more detailed info on how to upgrade between versions.
Django database backup¶
Alternatively, you can back up your database using Django’s
command. That way the backup is database agnostic and can be used in case you
want to change the database backend.
Prior to restoring the database you need to be running exactly the same Weblate
version the backup was made on. This is necessary as the database structure does
change between releases and you would end up corrupting the data in some way.
After installing the same version, run all database migrations using
Afterwards some entries will already be created in the database and you will have them in the database backup as well. The recommended approach is to delete such entries manually using the management shell (see Invoking management commands):
weblate shell >>> from weblate.auth.models import User >>> User.objects.get(username='anonymous').delete()
If you have enough backup space, simply back up the whole
is a safe bet even if it includes some files you don’t want.
The following sections describe what you should back up and what you
can skip in detail.
Dumped data for backups¶
Changed in version 4.7: The environment dump was added as
environment.yml to help in
restoring in the Docker environments.
Weblate dumps various data here, and you can include these files for more complete backups. The files are updated daily (requires a running Celery beats server, see Background tasks using Celery). Currently, this includes:
Weblate settings as
settings.py(there is also expanded version in
PostgreSQL database backup as
Environment dump as
The database backups are saved as plain text by default, but they can also be compressed
or entirely skipped using
To restore the database backup load it using dabase tools, for example:
psql --file=database.sql weblate
Version control repositories¶
The version control repositories contain a copy of your upstream repositories with Weblate changes. If you have Push on commit enabled for all your translation components, all Weblate changes are included upstream. No need to back up the repositories on the Weblate side as they can be cloned again from the upstream location(s) with no data loss.
SSH and GPG keys¶
If you are using SSH or GPG keys generated by Weblate, you should back up these locations. Otherwise you will lose the private keys and you will have to regenerate new ones.
User uploaded files¶
You should back up all user uploaded files (e.g. Visual context for strings).
The Celery task queue might contain some info, but is usually not needed for a backup. At most you will lose updates not yet been processed to translation memory. It is recommended to perform the fulltext or repository update upon restoration anyhow, so there is no problem in losing these.
Command line for manual backup¶
Using a cron job, you can set up a Bash command to be executed on a daily basis, for example:
$ XZ_OPT="-9" tar -Jcf ~/backup/weblate-backup-$(date -u +%Y-%m-%d_%H%M%S).xz backups vcs ssh home media fonts secret
The string between the quotes after XZ_OPT allows you to choose your xz options, for instance the amount of memory used for compression; see https://linux.die.net/man/1/xz
You can adjust the list of folders and files to your needs. To avoid saving the translation memory (in backups folder), you can use:
$ XZ_OPT="-9" tar -Jcf ~/backup/weblate-backup-$(date -u +%Y-%m-%d_%H%M%S).xz backups/database.sql backups/settings.py vcs ssh home media fonts secret
Restoring manual backup¶
Restore all data you have backed up.
Update all repositories using
weblate updategit --all