Guided setup¶
Use this path when you want to choose the right deployment model and avoid missing the setup decisions that matter later.
Step 1: Choose your deployment style¶
Local evaluation¶
Use this path when you want to:
- explore the admin UI
- verify local sign-in flows
- integrate a development application
- work with disposable or semi-persistent PostgreSQL state
Go to Local installation.
Production deployment¶
Use this path when you need:
- durable PostgreSQL storage
- real signing and encryption keys
- ingress or reverse proxy handling
- operational readiness, backups, and rollout discipline
Go to Production deployment.
Windows service deployment¶
Use this only when your environment requires a Windows-hosted API service and you do not want to run the container path.
Go to Windows service deployment.
Step 2: Confirm prerequisites¶
Local¶
- .NET 10 SDK
- Node.js LTS
- Git
Production¶
- Kubernetes cluster or equivalent container host
- PostgreSQL connectivity
- secret management process
- certificate and key ownership model
- ingress or reverse proxy plan
Step 3: Decide how keys and certificates are managed¶
Before production, decide:
- where OpenIddict signing material lives
- where OpenIddict encryption material lives
- how certificates are renewed and rolled out
- how applications validate issuer and JWKS metadata after rotation
The bundled Kubernetes path expects mounted PEM files projected from certificate secrets. See Certificates and keys.
Step 4: Plan the first bootstrap¶
For a clean first rollout, plan these items together:
- database connection string
- initial admin identity
- management web redirect URIs
- allowed browser origins
- first client application
- first service account, if machine-to-machine access is part of the rollout
The Kubernetes manifests support an optional admin seed secret for first-run onboarding.
Step 5: Validate the first successful flow¶
Your first validation should prove all of these:
- the API is healthy
- the management web can reach the API
- the first admin can sign in
- at least one client can complete its intended flow
- logs and health endpoints are visible to operators
From here, continue with Quick start for local usage or Operations overview for production rollout preparation.