Multi-Database and Schema Workflow

March 5, 2026 · View on GitHub

Use this tutorial when your application needs:

multiple databases (registry.extra), or
per-tenant/per-schema query context.

Choosing the Right Path

flowchart TD
    A["Need isolated data path"] --> B{"Different database server?"}
    B -- Yes --> C["Use registry.extra + using(database=...)"]
    B -- No --> D{"Same database, different schema?"}
    D -- Yes --> E["Use using(schema=...) / with_schema(...)"]
    D -- No --> F["Use default registry database/schema"]

Step 1: Declare Multiple Databases

{!> ../docs_src/registry/extra/declaration.py !}

Step 2: Write Data to an Extra Database

{!> ../docs_src/registry/extra/create.py !}

Step 3: Use Schema Context for Tenant-Like Routing

For schema-scoped operations:

from edgy.core.db import with_schema

with with_schema("tenant_a"):
    users = await User.query.all()

Or explicitly per query:

users = await User.query.using(schema="tenant_a").all()

Operational Notes

Keep registry lifecycle open (async with registry:) while serving requests.
Ensure target schemas exist before querying them.
Keep database names in using(database=...) aligned with registry.extra keys.

See Also