Edge SQL (D1)

Edge SQL gives your app a serverless SQLite database on Cloudflare D1. There is no server to size or keep awake: you pay for rows read, rows written, and storage. Use it for app data that fits SQLite well, such as settings, content, small catalogs, and per-tenant data, especially from Worker apps that cannot open a TCP connection to Postgres.

For a full Postgres, MySQL, or MongoDB server, see Postgres, MySQL & MongoDB.

Create a database#

From an app:

1. Open your app. On Overview, choose Add resource, then SQL database.
2. Keep Create new selected and enter a Name, such as main.
3. Choose Create, then redeploy the app.

To use a database your organization already has, choose Attach existing, pick it under Existing Database, and choose Attach.

The Databases page#

Your organization's databases are also listed under Databases in the top navigation (/projects/databases). From there:

1. Under New database, enter a name (lowercase letters, numbers, and dashes, up to 40 characters) and pick a location.
2. Choose Create.
3. Select the database. Under Attach to a project, choose the app, enter a binding name in UPPER_SNAKE_CASE (default DB), and choose Attach.
4. Redeploy that app.

Locations on the Databases page:

A location is a hint for where the primary copy lives. It cannot be changed later.

Limits#

Your plan sets how many Edge SQL databases your organization can have. Databases created from any app and from the Databases page count together.

At the limit, creating another shows "Your Pro plan includes 10 databases. Upgrade on the billing page for more."

Query from your code#

Worker apps (SSR and hybrid)#

The database is a D1 binding at env.NAME.

export default {
  async fetch(request, env) {
    const { results } = await env.MAIN
      .prepare('SELECT * FROM users WHERE id = ?')
      .bind(1)
      .all();
    return Response.json(results);
  },
};

Container apps#

POST the SQL and its parameters as JSON to http://{host}/query. Copy the exact host from the database's Connect tab. The answer is D1's {results, success, meta}. Each call runs one statement. Use ? placeholders and params for values.

use Illuminate\Support\Facades\Http;

$rows = Http::post('http://dply.my-app.main.internal/query', [
    'sql' => 'SELECT * FROM users WHERE id = ?',
    'params' => [1],
])->json('results');

const res = await fetch('http://dply.my-app.main.internal/query', {
  method: 'POST',
  headers: { 'content-type': 'application/json' },
  body: JSON.stringify({ sql: 'SELECT * FROM users WHERE id = ?', params: [1] }),
});
const { results } = await res.json();

Browse and query#

Choose Open on the database's card. The sheet has these tabs:

• Overview: size, number of tables, region (and the location you asked for), rows read and written this month, peak storage, and this month's cost.
• Connect: the code samples above with this database's host or binding filled in.
• Tables: lists tables. Select one to see its first 50 rows.
• Console: run SQL against the live database. Several statements separated by semicolons run in order, and the result shows the last one. Results show up to 200 rows.
• Settings: rename the binding (Save name). The app uses the new name after the next deploy; update your code to match. The database itself keeps its name.

The Databases page has the same kind of console under SQL, with Run.

Pricing#

Edge SQL bills every unit from the first:

Storage is billed as the database's largest size during the month, for the full month. Usage is collected daily, so the sheet's figure runs through yesterday. This bills at the invoice line SQL, queues and key-value.

Delete a database#

From the app, choose Delete on the card, or Delete database under Settings. This deletes the database and every row in it, and removes it from the app. It cannot be undone. If the app's live deploy still uses the database, dply detaches it now and deletes it after your next deploy. See Detach or delete.

On the Databases page, select the database, type its name to confirm, and choose Delete database.

Related#

• Resources overview
• Postgres, MySQL & MongoDB
• Database pools
• Usage & metering