Supabase backup retention planner
A no-login worksheet for deciding whether daily dumps, PITR, physical/base backups, Storage object versioning, or a managed Postgres move fits your recovery target before a costly backup surprise.
1
PITR solves a different problemPoint-in-time recovery covers small recovery windows with fine granularity; it is not the same as cheap long-retention daily archives.
2
pg_dump needs a restore drillA dump is not a backup plan until a disposable restore proves recovery time, role behavior, extensions, and app smoke tests.
3
Database backups miss object bytesSupabase database backups include Storage metadata, not deleted object contents. Plan bucket/object recovery separately.
Runs locally in your browser. Paste only redacted notes. Do not paste database URLs, passwords, connection strings, JWT secrets, service-role keys, customer rows, private screenshots, payment data, full names, private handles, or payment identifiers.
Ready.
Backup review packet
What this checks
- Whether long retention is a daily archive problem, a PITR problem, or both.
- Whether pg_dump is backed by a restore drill, not just a scheduled export.
- Whether Storage object bytes have a backup/versioning path separate from database metadata.
- Whether custom roles, extensions, migration replay, and app smoke tests are captured after restore.
- Whether owner, alerting, retention, egress, and restore-time expectations are written down.
Source-backed problem set
Supabase backups
Daily backups, PITR restore flow, restore downtime, Storage-object exclusion, and role-password caveats.
Logical backup download
PITR or physical backups can remove the downloadable logical backup file; use CLI or dump when a logical export is needed.
Self-hosting responsibilities
Managed backups and PITR are unavailable in self-hosted Supabase, and disaster recovery becomes your responsibility.
Postgres PITR
Point-in-time recovery requires a base backup plus continuous archived WAL files, not only logical dumps.