Projects
A Flux project is the deployable backend unit: database surface, REST API surface, routing identity, authentication configuration, and engine assignment (v1 dedicated vs v2 shared).
What you will learn
- Project vs database vs API
- What operators manage through CLI/dashboard
- Where to go for isolation details
The idea
Calling a project “a database” understates what Flux provisions. The control plane tracks metadata, URLs, secrets, and lifecycle while the data plane serves queries. Slug and hash appear in hostnames; internal identifiers (like tenant_id on v2) are immutable even if marketing slugs change.
How it works
Operators typically:
- Create / destroy projects
- Run migrations (
flux push) - Rotate secrets (carefully—see Project secrets)
- Start / stop / repair stacks
Developers consume the Service URL and credentials appropriate to their engine.
Example
bash
flux create "billing-svc"flux list# Later: flux push … --project <slug> --hash <hash> # values from flux list
Output includes identifiers you use in URLs and automation—not interchangeable with “just a Postgres URL” unless you bypass PostgREST entirely (not the default product path).