Backup and Recovery
This guide covers what to back up for a Simba Intelligence deployment and basic recovery planning. Simba Intelligence is deployed as part of the Self-Service Analytics (formerly Logi Composer) Helm chart, so its data sits alongside the Self-Service Analytics data.Production Note: For production, use an external managed database service (AWS RDS, Azure Database, Google Cloud SQL) rather than the chart’s in-cluster PostgreSQL, and back it up with that service’s own tooling.
What to Back Up
🎯 Critical Data:- Databases — a single PostgreSQL instance holds the
simbaintelligencedatabase alongside the Self-Service Analytics databases (zoomdata,zoomdata-upload,zoomdata-keyset,zoomdata-user-auditing,zoomdata-qe) - Persistent volumes — see below
- Your Helm values file — keep it in version control; it is what you reinstall from
Database Backups
Back up all of the databases listed above together — Simba Intelligence and Self-Service Analytics share a deployment and should be recovered to a consistent point in time. Recommended schedule: Daily automated backups with 30-day retentionPersistent Storage Backups
Back up the PersistentVolumeClaims the chart creates. To see exactly which ones your configuration produces:Recovery Process
- Scale the Deployments and StatefulSets to zero replicas
- Restore the databases and persistent volumes from your most recent backup
- Verify the restored data
- Reapply the Helm chart with your saved values file
Best Practices
Schedule:- Daily database backups
- Weekly storage backups
- Monthly recovery testing
- Store backups separately from your main system
- Encrypt sensitive database backups
- Limit access to backup files
- Regularly test your restore process
- Verify backup integrity
- Practice recovery during maintenance windows
- Keep 30 days of daily backups
- Keep 12 weeks of weekly backups
- Store critical backups offsite
See also: Upgrade Procedures for pre-upgrade backup planning.