Postgres Ops

Keep a production Postgres safe to lose and fast to serve: backups you have actually restored, recovery you have actually tested, and pooling that survives serverless connection storms. Restore and recovery operations can overwrite data — they run behind explicit confirmation, against the right target.

Contract

Inputs:

• A Postgres instance (managed or self-managed); mode backup / restore / pooling / dr.

Outputs:

• A backup/PITR plan or verification, a restore result, a pooling configuration, or a DR runbook.