Use Neon Postgres through Drizzle ORM

Neon is a fully managed serverless Postgres, separating compute and storage to offer features like autoscaling, branching and bottomless storage. Neon can be used from Bun directly using the @neondatabase/serverless driver or through an ORM like Drizzle. Drizzle ORM supports both a SQL-like “query builder” API and an ORM-like Queries API. Get started by creating a project directory, initializing the directory using bun init, and installing Drizzle and the Neon serverless driver.

terminal

mkdir bun-drizzle-neon
cd bun-drizzle-neon
bun init -y
bun add drizzle-orm @neondatabase/serverless
bun add -D drizzle-kit

Create a .env.local file and add your Neon Postgres connection string to it.

.env.local

DATABASE_URL=postgresql://usertitle:password@ep-adj-noun-guid.us-east-1.aws.neon.tech/neondb?sslmode=require

We will connect to the Neon database using the Neon serverless driver, wrapped in a Drizzle database instance.

db.ts

import { neon } from "@neondatabase/serverless";
import { drizzle } from "drizzle-orm/neon-http";

// Bun automatically loads the DATABASE_URL from .env.local
// Refer to: https://bun.com/docs/runtime/environment-variables for more information
const sql = neon(process.env.DATABASE_URL!);

export const db = drizzle(sql);

To see the database in action, add these lines to index.ts.

index.ts

import { db } from "./db";
import { sql } from "drizzle-orm";

const query = sql`select 'hello world' as text`;
const result = await db.execute(query);
console.log(result.rows);

Then run index.ts with Bun.

terminal

bun run index.ts

[
  {
    text: "hello world",
  }
]

We can define a schema for our database using Drizzle ORM primitives. Create a schema.ts file and add this code.

schema.ts

import { pgTable, integer, serial, text, timestamp } from "drizzle-orm/pg-core";

export const authors = pgTable("authors", {
  id: serial("id").primaryKey(),
  title: text("name").notNull(),
  bio: text("bio"),
  createdAt: timestamp("created_at").notNull().defaultNow(),
});

We then use the drizzle-kit CLI to generate an initial SQL migration.

bunx drizzle-kit generate --dialect postgresql --schema ./schema.ts --out ./drizzle

This creates a new drizzle directory containing a .sql migration file and meta directory.

File Tree

drizzle
├── 0000_aspiring_post.sql
└── meta
    ├── 0000_snapshot.json
    └── _journal.json

We can execute these migrations with a migrate.ts script. This script creates a new connection to the Neon database and executes all unexecuted migrations in the drizzle directory.

migrate.ts

import { db } from "./db";
import { migrate } from "drizzle-orm/neon-http/migrator";

const main = async () => {
  try {
    await migrate(db, { migrationsFolder: "drizzle" });
    console.log("Migration completed");
  } catch (error) {
    console.error("Error during migration:", error);
    process.exit(1);
  }
};

main();

We can run this script with bun to execute the migration.

terminal

bun run migrate.ts

Migration completed

We can now add some data to our database. Create a seed.ts file with the following contents.

seed.ts

import { db } from "./db";
import * as schema from "./schema";

async function seed() {
  await db.insert(schema.authors).values([
    {
      title: "J.R.R. Tolkien",
      bio: "The creator of Middle-earth and author of The Lord of the Rings.",
    },
    {
      title: "George R.R. Martin",
      bio: "The author of the epic fantasy series A Song of Ice and Fire.",
    },
    {
      title: "J.K. Rowling",
      bio: "The creator of the Harry Potter series.",
    },
  ]);
}

async function main() {
  try {
    await seed();
    console.log("Seeding completed");
  } catch (error) {
    console.error("Error during seeding:", error);
    process.exit(1);
  }
}

main();

Then run this file.

terminal

bun run seed.ts

Seeding completed

We now have a database with a schema and sample data. We can use Drizzle to query it. Replace the contents of index.ts with the following.

index.ts

import * as schema from "./schema";
import { db } from "./db";

const result = await db.select().from(schema.authors);
console.log(result);

Then run the file. You should see the three authors we inserted.

terminal

bun run index.ts

[
  {
    id: 1,
    title: "J.R.R. Tolkien",
    bio: "The creator of Middle-earth and author of The Lord of the Rings.",
    createdAt: 2024-05-11T10:28:46.029Z,
  }, {
    id: 2,
    title: "George R.R. Martin",
    bio: "The author of the epic fantasy series A Song of Ice and Fire.",
    createdAt: 2024-05-11T10:28:46.029Z,
  }, {
    id: 3,
    title: "J.K. Rowling",
    bio: "The creator of the Harry Potter series.",
    createdAt: 2024-05-11T10:28:46.029Z,
  }
]

This example used the Neon serverless driver’s SQL-over-HTTP functionality. Neon’s serverless driver also exposes Client and Pool constructors to enable sessions, interactive transactions, and node-postgres compatibility. Refer to Neon’s documentation for a complete overview. Refer to the Drizzle website for more documentation on using the Drizzle ORM.

Overview

Deployment

Runtime & Debugging

Utilities

Ecosystem & Frameworks

HTTP & Networking

WebSocket

Processes & System

Package Manager

Test Runner

Runtime & Debugging

Module System

File System

Utilities

HTML Processing

Binary Data

Streams

Use Neon Postgres through Drizzle ORM