Backups and upgrades

Backups

All data is in the SQLite database (/data/seatbit.db) and the uploaded avatars (/data/uploads).

The image contains a backup script that makes a consistent copy while Seatbit runs:

docker exec seatbit node scripts/backup.mjs

It writes seatbit-YYYYMMDD-HHMMSS.db to BACKUP_DIR (default /backups) and deletes backups older than BACKUP_KEEP_DAYS (default 14). Run it daily from cron (or a Kubernetes CronJob mounting the same volume) and copy the backups off the server.

To restore, stop Seatbit, replace /data/seatbit.db with a backup (remove seatbit.db-wal and seatbit.db-shm next to it) and start it again.

Upgrades

Pull the new image and restart the container. Migrations run on startup. Take a backup first.

docker compose pull && docker compose up -d

There is one database, so run a single instance and stop the old container before the new one starts (Recreate strategy on Kubernetes).

Logs and health

• GET /api/health answers {"ok":true} while the server and database work.
• Errors, sent reminders and (without SMTP) emails go to the container log.