PostgreSQL
As a Project administrator or developer, use these guides to create and operate a managed PostgreSQL database. Select a published plan for the version, capacity, availability, and recovery support your application needs.
Create and connect
Start with the creation and connection procedures:
- Create a PostgreSQL service with a manifest, or use the console.
- Connect applications through a binding with verified TLS.
- Configure credentials and rotation for declared roles or supported existing SQL logins.
PostgreSQL provides a primary endpoint, replica endpoints, and optional PgBouncer pooling. Endpoint selection and credential roles are independent. See the connection guide for availability checks and client examples.
Operate and recover
Choose the guide for your task:
| Task | Guide |
|---|---|
| Scale replicas, resize compute, expand storage, change settings, or upgrade | PostgreSQL operations |
| Schedule backups, restore into another service, restore in place, or choose a recovery time | PostgreSQL backups and recovery |
| Connect from outside the Project through a supported Gateway | PostgreSQL external access |
| Remove a database or inspect retained resources | PostgreSQL deletion |
Plan entitlement and service state determine which actions you can request. Test recovery before you rely on backups, and check the documented limits for each operation.
Shared managed service tasks
The same platform rules apply to PostgreSQL and other service families:
- Classes and plans define available capacity and operations.
- Request an operation explains approval, execution windows, and results.
- Check service status explains readiness, stale observations, and refused changes.
- Delete a managed service explains protection and Project deletion.