Database Management (Drizzle ORM + Cloudflare D1)#

This project uses Drizzle ORM for database schema definition and migrations, and Cloudflare D1 for the database service.

Migrations#

Migrations allow for controlled updates to the database schema.

1. Define Schema#

Modify the schema definitions in src/lib/server/schema.ts.

2. Generate Migration Files#

After changing the schema, generate the SQL migration files:

bunx drizzle-kit generate

This command compares your schema file to the database state (tracked locally) and creates a new .sql file in the drizzle/ directory.

3. Apply Migrations Locally#

To apply pending migrations to your local development database (defined in wrangler.json and managed by Miniflare), run:

bunx drizzle-kit migrate

Note: Sometimes, if the local .wrangler/state gets out of sync, you might need to delete the .wrangler/state directory and run the migrate command again. (Drizzle Kit requires @libsql/client or better-sqlite3 installed as a dev dependency for local migrations).

4. Apply Migrations Remotely (Cloudflare D1)#

To apply pending migrations to the live Cloudflare D1 database, use the wrangler CLI. Make sure your wrangler.json specifies the correct database_name.

For the development database (ananas-auth-dev):

bunx wrangler d1 migrations apply ananas-auth-dev --remote

Replace ananas-auth-dev with your production database name if you have a separate one.

Note: If you encounter errors like "table already exists" during remote migration, it might mean a previous migration was applied manually or failed partway. You may need to manually adjust the state or temporarily comment out statements in the corresponding migration file before retrying the apply --remote command.