Skip to content

FAQ

General

Q: Does this work with Odoo Community Edition?

Yes. Both backup methods work with CE and EE. The Odoo Endpoint method uses /web/database/backup, which is available in all editions. The pg_dump method bypasses Odoo entirely and connects directly to PostgreSQL.

Q: What versions of Odoo are supported?

The Odoo Endpoint method has been tested with Odoo 14 through 17. The pg_dump method supports any PostgreSQL-backed Odoo version (practically all modern versions) and does not depend on the Odoo application at all.

Q: Can I back up multiple Odoo instances from one orchestrator?

Yes. There is no hard limit on the number of instances. Each instance has its own jobs, schedules, and notification bindings. All backups land in separate subdirectories under /data/backups/.

Q: Does the backup include the filestore (attachments)?

  • Odoo Endpoint: Yes, always. The /web/database/backup endpoint returns a ZIP that includes both the database dump and the filestore.
  • pg_dump: Only the database. Filestore sync is on the roadmap but not yet implemented.

Q: My Odoo instance is behind a self-signed certificate. Will this work?

No — the orchestrator enforces TLS certificate validation and does not support verify=False. Use a certificate signed by a recognized CA (including Let's Encrypt), or add your CA certificate to the orchestrator's trust store via the SSL_CERT_FILE environment variable.

Backup Methods

Q: Which backup method should I use?

Use Odoo Endpoint if: - You don't have direct database access - You want the simplest setup (just the master password) - You need the filestore included automatically

Use pg_dump if: - The Odoo master password isn't available - You have very large databases (pg_dump is faster for large DBs) - You need fine-grained control (pg_dump custom format is more portable)

Q: What's the Odoo master password?

It's the admin_passwd in your Odoo odoo.conf. It's required to authenticate to the database manager endpoints. Find it with:

grep admin_passwd /etc/odoo/odoo.conf

Q: The test connection fails for pg_dump. What should I check?

  1. Is the PostgreSQL host reachable from the orchestrator container? Try ping or nc -z <host> 5432.
  2. Does the PostgreSQL user have CONNECT and SELECT privileges on the target database?
  3. Is pg_hba.conf configured to allow connections from the orchestrator's IP?
  4. Is the port correct? The default is 5432.

Scheduling

Q: Jobs aren't firing. How do I debug?

  1. Check that the beat and worker containers are running: docker compose ps.
  2. Check the beat logs for schedule sync messages: docker compose logs beat.
  3. Check the worker logs for task execution: docker compose logs worker.
  4. Verify the job has Enabled toggled on in the UI.
  5. Check that Redis is healthy: docker compose exec redis redis-cli ping should return PONG.

Q: I changed a cron expression but the job still fires on the old schedule. Is there a delay?

The refresh_schedules task runs every 60 seconds. Wait up to 60 seconds after saving a change for the new schedule to take effect. You can also manually trigger a refresh by restarting the beat container: docker compose restart beat.

Q: Can I run a job more frequently than once per minute?

No. Standard cron syntax has a 1-minute minimum granularity. For sub-minute intervals, you would need to implement a custom task loop — this is outside the scope of the current feature set.

Notifications

Q: Emails are going to spam. How do I fix this?

  1. Set up SPF and DKIM records for your sending domain.
  2. Use a dedicated transactional email service (Mailgun, SendGrid, AWS SES) rather than your personal Gmail.
  3. Configure the From address to match the sending domain (not a generic gmail.com address).
  4. Avoid spam-trigger words in the subject (the current templates are conservative).

Q: Can I send notifications to a Slack channel?

Not directly — Slack is not a built-in channel type. You can work around this with a Telegram → Slack bridge, or by creating a Telegram channel pointing to a Slack-integrated Telegram bot. Native Slack support is on the roadmap.

Cloud Sync

Q: Can I sync to an S3-compatible storage (MinIO, Backblaze B2, etc.)?

Not yet. S3 support is planned. Currently only Google Drive, Dropbox, and OneDrive are supported.

Q: How large a file can be uploaded to Google Drive?

Google Drive's API supports uploads up to 5 TB for service accounts. For most Odoo deployments, this is not a practical limit.

Operations

Q: How do I back up the orchestrator itself?

Back up the /data/ directory. It contains:

  • db/app.sqlite — the SQLite database (all configuration, job history, audit log)
  • secret.key — the Fernet encryption key
  • backups/ — the backup archives themselves

A simple approach:

docker compose stop
tar czf odoo-backup-app-$(date +%Y%m%d).tar.gz /path/to/data/
docker compose start

Or use VACUUM INTO for a hot copy of the SQLite database without stopping services.

Q: How do I upgrade to a new version?

docker compose pull
docker compose up -d

Alembic migrations run automatically on startup. The database schema is upgraded before the application accepts requests.

Q: I lost my admin password and didn't set up email recovery. Am I locked out?

Yes — without password recovery configured, there is no automated way to reset the password. Recovery options:

  1. Direct database edit: Stop the app, use sqlite3 /data/db/app.sqlite, update the password_hash column in the users table with a new Argon2 hash. Restart.
  2. Admin CLI (not yet implemented): A manage.py reset-password command is planned.

This is why the setup checklist prominently highlights configuring password recovery.

Q: Can I run the application without Docker?

Yes. Install the Python dependencies from requirements.txt, run Redis separately, and start the three processes manually:

uvicorn backend.app.main:app --port 8080
celery -A backend.app.tasks.celery_app worker -Q default,backups,notifications,cloud
celery -A backend.app.tasks.celery_app beat

Set the OB_* environment variables as needed. The frontend must be pre-built (npm run build in frontend/) or served by a separate dev server.

Q: Is there a REST API I can call from other tools?

Yes. The full API is available at /api/. It uses the same session cookie authentication as the browser. You can log in with a POST to /api/auth/login and then use the returned cookie in subsequent requests. API documentation (OpenAPI) is available at /api/docs when the app is running.