SpiderFolio
en
Docs

Self-hosting

Backups

Back up with backup.sh, schedule it with cron, keep copies off the server and restore with restore.sh.

A complete SpiderFolio backup has three parts:

  • The database: pages, permissions, history, audit log… everything that isn’t a file.
  • The files: the files volume (or your bucket, if you use S3).
  • The .env file: without it, two-step verification secrets and each company’s own SMTP passwords can’t be decrypted.

deploy/backup.sh saves all three and deploy/restore.sh brings them back.

Take a backup

From the deploy/ folder (where compose.yaml lives), with the services running:

./backup.sh /var/backups/spiderfolio

It creates a folder named after the UTC date and time (for example 20261026-033000/) with:

File Contents
db.dump The full database (pg_dump in custom format)
files.tar.gz The contents of the files volume
env A copy of .env, with 600 permissions
  • Without a destination folder, it uses ./backups.
  • It keeps the last 14 backups and deletes older ones. To change that: KEEP=30 ./backup.sh ….
  • The app keeps working while the backup runs.

Schedule a daily backup

With crontab -e, as the user that runs Docker:

30 3 * * * cd /opt/spiderfolio/deploy && ./backup.sh /var/backups/spiderfolio >> /var/log/spiderfolio-backup.log 2>&1

Replace /opt/spiderfolio/deploy with the folder where you cloned the project. Check the log now and then: each backup ends with “Copia lista en …” (backup ready in …).

Keep copies off the server

A backup on the same disk doesn’t protect you if you lose the server. Copy them elsewhere: another server, another provider or a bucket. For example, after the backup:

0 4 * * * rsync -a --delete /var/backups/spiderfolio/ backups@other-server:/backups/spiderfolio/

Or with rclone to a bucket. Backups include .env with every secret: store them encrypted or in a destination with restricted access.

If you use S3

With STORAGE_DRIVER=s3, files.tar.gz is empty: the files aren’t on the server. Protect the bucket with your provider’s tools: turn on versioning or replication to another bucket. The database and .env are still backed up by backup.sh.

Restore

Restoring replaces the current database and files with those from the backup.

cd /opt/spiderfolio/deploy
docker compose up -d                                  # the services must exist
./restore.sh /var/backups/spiderfolio/20261026-033000

The script asks you to type RESTAURAR to continue. It then stops the app and the worker, recreates the database, replaces the files and starts everything again. When it’s done, sign in to the app and check the audit log.

On a new server

  1. Install Docker and clone the project as in Install on your server.
  2. Copy the backup’s env as deploy/.env (cp …/env .env && chmod 600 .env).
  3. Start the services: docker compose up -d (with the same profiles you used before).
  4. Restore: ./restore.sh /path/to/backup.
  5. If the server’s address changed, update DNS and, if needed, APP_URL and DOMAIN.

If the backup comes from an older version, the app applies pending migrations on startup and brings it up to date automatically.

Test your restores

Restore a backup on a separate machine from time to time, for example once a month. A backup that has never been restored isn’t a backup.