Belajar Drizzle ORM - Project Structure & Scalability

#Pendahuluan

Setelah di episode 18 kita mempelajari Testing dengan In-Memory SQLite — unit testing dan integration testing — pada episode ini kita membahas project structure: bagaimana mengorganisasi file dan folder dalam proyek Drizzle ORM agar scalable, maintainable, dan mudah dipahami oleh tim.

Project structure yang baik mengurangi technical debt dan memudahkan onboarding developer baru.

#Struktur Folder untuk Proyek Kecil-Sedang

plaintext

src/
  db/
    schema/          # Drizzle schema definitions
      users.ts
      posts.ts
      relations.ts   # Centralized relations
    index.ts         # db instance export
    migrations/      # Generated by drizzle-kit
  services/          # Business logic, queries
    user.service.ts
    post.service.ts
  routes/            # API routes
    user.routes.ts
    post.routes.ts
drizzle.config.ts

#Penjelasan

db/schema/: satu file per tabel/domain. Memudahkan navigasi.
db/index.ts: export db instance. Import dari satu tempat.
services/: isolasi queries dari routes. Reusable query functions.
routes/: API routes. Memanggil service, bukan query langsung.

#Modular Schema

#Satu File per Domain

src/db/schema/users.ts

import { pgTable, serial, text, integer } from 'drizzle-orm/pg-core'
 
export const users = pgTable('users', {
  id: serial('id').primaryKey(),
  name: text('name').notNull(),
  email: text('email').notNull().unique(),
})

src/db/schema/posts.ts

import { pgTable, serial, text, integer } from 'drizzle-orm/pg-core'
 
export const posts = pgTable('posts', {
  id: serial('id').primaryKey(),
  title: text('title').notNull(),
  authorId: integer('author_id').references(() => users.id),
})

#Centralized Relations

src/db/schema/relations.ts

import { relations } from 'drizzle-orm'
import { users } from './users'
import { posts } from './posts'
 
export const usersRelations = relations(users, ({ many }) => ({
  posts: many(posts),
}))
 
export const postsRelations = relations(posts, ({ one }) => ({
  author: one(users, { fields: [posts.authorId], references: [users.id] }),
}))

#Query Service Pattern

#Isolasi Queries dari Routes

src/services/user.service.ts

import { db } from '../db'
import { users } from '../db/schema'
import { eq, and } from 'drizzle-orm'
 
export async function getAllUsers() {
  return db.select().from(users)
}
 
export async function getUserById(id: number) {
  return db.select().from(users).where(eq(users.id, id))
}
 
export async function createUser(data: InsertUser) {
  const [user] = await db.insert(users).values(data).returning()
  return user
}

#Routes Memanggil Service

src/routes/user.routes.ts

import { getAllUsers, getUserById, createUser } from '../services/user.service'
 
app.get('/users', async () => {
  const users = await getAllUsers()
  return users
})
 
app.get('/users/:id', async ({ params }) => {
  const user = await getUserById(Number(params.id))
  return user
})

Tip

Query service pattern memungkinkan kalian mengganti implementation tanpa mengubah routes. Jika perlu migrate dari Drizzle ke ORM lain, cukup ubari service layer — routes tidak terpengaruh.

#Scaling ke Microservices

#Shared Schema Package

Untuk microservices yang berbagi database, buat schema sebagai npm package terpisah:

plaintext

packages/
  shared-schema/     # npm package
    src/
      users.ts
      posts.ts
    package.json
  api-service/       # Microservice 1
    src/
      routes/
    package.json
  web-service/       # Microservice 2
    src/
      routes/
    package.json

#Database per Service

Microservices pattern merekomendasikan database per service — setiap service punya database sendiri. Schema pattern yang sama bisa digunakan di setiap service.

#Penutup

Pada episode 19 ini, kalian telah memahami Project Structure & Scalability:

Modular schema: satu file per domain, centralized relations.
Query service pattern: isolasi queries dari routes.
Struktur folder: db/schema/, services/, routes/.
Scaling: shared schema package untuk microservices.
Database per service: pattern microservices.

Di episode 20 selanjutnya — episode terakhir — kita akan mempelajari Production Deployment & Best Practice — database connection management, migration strategy, performance monitoring, security, CI/CD integration, dan health check endpoint. Sampai jumpa di episode 20!