setup-database

Database Setup Wizard

I'll set up database storage for this Worker: add the binding, regenerate types, and create starter schema and example code you can adapt. Wrangler runs as pnpm exec wrangler ....

Step 1: Choose a store

D1 - serverless SQLite. Relational data, JOINs, transactions (via batch). Users, content, orders, reporting. Docs: https://developers.cloudflare.com/d1/

KV - global key-value store, eventually consistent, built for many reads and few writes. Sessions, caching, feature flags, configuration. Docs: https://developers.cloudflare.com/kv/

Need per-entity strong consistency or real-time coordination (counters, rooms, rate limiters that must be exact)? That is a Durable Object - use /add-binding. Files belong in R2, also via /add-binding.

I'll ask:

1. D1 or KV?
2. The resource name (D1: lowercase-with-hyphens, e.g. my-app-db; KV binding: UPPER_SNAKE_CASE, e.g. SESSIONS).
3. Does the data have to stay in a region? If so, which jurisdiction: eu, us or fedramp. This can only be set when the resource is created.
4. (Optional) What it is for, so the examples fit.

Before editing, I check wrangler.jsonc for an existing binding with the same name.

---

D1 flow

1. Add the binding

Two routes - I'll recommend the first unless a jurisdiction or location hint is needed:

Automatic provisioning. Add the binding without an ID. pnpm dev uses a local database, and the first wrangler deploy creates the real one and writes database_id back into wrangler.jsonc (commit that change).

"d1_databases": [{ "binding": "DB", "database_name": "my-app-db" }]

Create it now. Wrangler creates the database and updates the config itself:

pnpm exec wrangler d1 create my-app-db --binding DB --update-config
# Optional: --jurisdiction eu|us|fedramp   (data localisation, fixed at creation)
# Optional: --location weur|eeur|apac|oc|wnam|enam   (primary location hint)

Docs: https://developers.cloudflare.com/d1/configuration/data-location/