非同期データベースアクセス
`asyncpg`や`SQLModel`などの非同期データベースドライバーとORMを使った、ノンブロッキングなデータベース操作を学びます。
「非同期データベースアクセス」はCoddyKit上の無料FastAPI Backend Development Bootcampレッスンです。 これはレッスン2/4です。 下記で完全なレッスンを無料で読むことができます。その後、ブラウザ内の組み込みコードエディタと24時間対応のAIチューターでハンズオン演習できます。 これはFastAPI Backend Development Bootcamp学習パスの一部であり、ウェブとCoddyKitアプリ全体で進捗が同期されます。 FastAPI Backend Development Bootcampコースには全4レッスンが含まれています。
このレッスンの一部はまだ翻訳されておらず、英語で表示されています。
Why Async DB Access?
When building high-performance web applications, especially with frameworks like FastAPI, database operations can often become a bottleneck. Traditional database calls are "blocking", meaning your application waits for the database to respond before moving on.
Asynchronous database access allows your application to perform other tasks while waiting for the database, preventing your API from becoming unresponsive under heavy load. This is crucial for scalability!
Sync vs. Async DB Calls
Imagine a restaurant where one chef handles everything. If a customer orders a complex dish (a database query), the chef stops all other work until that dish is complete. This is synchronous I/O.
In an asynchronous setup, the chef starts the complex dish, then immediately moves to prepare simpler dishes or take new orders while the complex dish cooks in the background. When the complex dish is ready, the chef picks it up. This non-blocking approach boosts efficiency!
Meet asyncpg: The Async Driver
For PostgreSQL, the go-to asynchronous driver in Python is asyncpg. It's a high-performance library specifically designed to work with Python's asyncio framework.
- Fast: Written partly in C for speed.
- Asynchronous: Fully non-blocking.
- Direct: Provides a direct interface to PostgreSQL.
It's often used as the underlying driver for async ORMs or when you need fine-grained control.
asyncpg in Action
Let's see a basic example of connecting to a PostgreSQL database and running a simple query using asyncpg. Remember to replace placeholder credentials with your own!
import asyncio
import asyncpg
async def main():
conn = None
try:
conn = await asyncpg.connect(user='postgres', password='mysecretpassword',
database='testdb', host='localhost')
print("Connected to PostgreSQL!")
# Execute a query
result = await conn.fetchval('SELECT 1 + 1')
print(f"Query result: {result}") # Should be 2
except Exception as e:
print(f"Error: {e}")
finally:
if conn:
await conn.close()
print("Connection closed.")
if __name__ == "__main__":
asyncio.run(main())Awaiting Database Calls
In the previous example, you saw the await keyword before asyncpg.connect() and conn.fetchval(). This is crucial for asynchronous operations.
awaittells Python: "This operation might take time, so pause here and let other tasks run in the meantime."- When the database operation completes, the task resumes from where it left off.
- This non-blocking wait is what makes your FastAPI application scalable.
SQLModel: Async ORM Power
While asyncpg gives you low-level control, an Object Relational Mapper (ORM) simplifies database interactions by mapping database tables to Python objects. SQLModel is a modern, async-first ORM built on:
- Pydantic: For data validation and serialization.
- SQLAlchemy: The powerful and mature Python SQL toolkit.
It lets you define models once and use them for both your API request/response and database schema!
Setting up SQLModel for Async
To use SQLModel asynchronously, you need an asynchronous database engine. This typically involves using an async driver like asyncpg (which SQLAlchemy can use via asyncio). Here's how you'd set up the engine:
from sqlmodel import create_engine, SQLModel
import asyncio
# Replace with your actual async PostgreSQL connection string
# The 'postgresql+asyncpg' part tells SQLAlchemy to use asyncpg
DATABASE_URL = "postgresql+asyncpg://postgres:mysecretpassword@localhost/testdb"
async def main():
engine = create_engine(DATABASE_URL, echo=True)
print("Async SQLModel engine created.")
# In a real app, you'd usually create tables here
# async with engine.begin() as conn:
# await conn.run_sync(SQLModel.metadata.create_all)
# Just demonstrating engine creation for this example
await engine.dispose()
print("Engine disposed.")
if __name__ == "__main__":
asyncio.run(main())Your First SQLModel
Defining a model in SQLModel is super intuitive. You inherit from SQLModel and use Pydantic-like field declarations. This single definition creates both your database table schema and your API's data validation schema!
from typing import Optional
from sqlmodel import Field, SQLModel
class Hero(SQLModel, table=True):
id: Optional[int] = Field(default=None, primary_key=True)
name: str = Field(index=True)
secret_name: str
age: Optional[int] = Field(default=None, index=True)
# This model can now be used with FastAPI for request bodies
# and with SQLAlchemy for database interactions.
print("Hero model defined successfully!")Async CRUD with SQLModel
Now let's perform a simple Create and Read operation using our Hero model and the async engine. We'll use AsyncSession from SQLAlchemy's ORM for database interactions.
from typing import Optional
from sqlmodel import Field, SQLModel, create_engine, Session, select
from sqlalchemy.ext.asyncio import AsyncSession, create_async_engine
from sqlalchemy.orm import sessionmaker
import asyncio
DATABASE_URL = "postgresql+asyncpg://postgres:mysecretpassword@localhost/testdb"
async_engine = create_async_engine(DATABASE_URL, echo=False)
AsyncSessionLocal = sessionmaker(autocommit=False, autoflush=False, bind=async_engine, class_=AsyncSession)
class Hero(SQLModel, table=True):
id: Optional[int] = Field(default=None, primary_key=True)
name: str = Field(index=True)
secret_name: str
age: Optional[int] = Field(default=None, index=True)
async def create_db_and_tables():
async with async_engine.begin() as conn:
await conn.run_sync(SQLModel.metadata.create_all)
print("Database tables created/updated.")
async def create_hero(hero: Hero):
async with AsyncSessionLocal() as session:
session.add(hero)
await session.commit()
await session.refresh(hero)
print(f"Created hero: {hero.name} (ID: {hero.id})")
return hero
async def get_heroes():
async with AsyncSessionLocal() as session:
statement = select(Hero)
results = await session.execute(statement)
heroes = results.scalars().all()
print("\nAll Heroes:")
for hero in heroes:
print(f"- {hero.name} (Age: {hero.age})")
return heroes
async def main():
await create_db_and_tables()
hero_1 = Hero(name="Deadpond", secret_name="Dive Wilson", age=28)
hero_2 = Hero(name="Spider-Boy", secret_name="Pedro Parqueador")
hero_3 = Hero(name="Rusty-Man", secret_name="Tommy Sharp", age=48)
await create_hero(hero_1)
await create_hero(hero_2)
await create_hero(hero_3)
await get_heroes()
await async_engine.dispose()
if __name__ == "__main__":
asyncio.run(main())Async DB Check
You've learned about the importance of asynchronous database access and explored tools like asyncpg and SQLModel. Let's test your understanding!
Recap: Async DB for Scale
Great job! You've successfully explored asynchronous database access.
- We understood why non-blocking I/O is vital for high-performance FastAPI apps.
- We introduced
asyncpgas a low-level async PostgreSQL driver. - We learned about
SQLModel, an async-first ORM combining Pydantic and SQLAlchemy. - We saw practical examples of setting up and performing CRUD operations with these tools.
Mastering async database interactions is a key step towards building truly scalable and responsive backend services!
よくある質問
「非同期データベースアクセス」レッスンは無料ですか?
はい。「非同期データベースアクセス」の完全なテキストはこのウェブで無料で読めます。インタラクティブに演習し(組み込みコードエディタと24時間対応のAIチューター)、FastAPI Backend Development Bootcampコースの残りをアンロックするには、CoddyKit PROにアップグレードしてください。 FastAPI Backend Development Bootcampコースには全4レッスンが含まれています。
「非同期データベースアクセス」で何を学びますか?
`asyncpg`や`SQLModel`などの非同期データベースドライバーとORMを使った、ノンブロッキングなデータベース操作を学びます。 ブラウザで直接実行するハンズオンコードでFastAPI Backend Development Bootcampを演習し、24時間対応のAIチューターがレッスンを進める中での質問に答えます。
FastAPI Backend Development Bootcampを始めるのに経験は必要ですか?
事前経験は必要ありません。CoddyKitのFastAPI Backend Development Bootcampは初級者から上級者向けに構成されているため、ここから始めるか最初から始めて、自分のペースで進むことができます。 これはレッスン2/4です。
「非同期データベースアクセス」レッスンにはどのくらい時間がかかりますか?
ほとんどのCoddyKitレッスンは約5~10分かかります。各レッスンはコンパクトでインタラクティブなので、着実に進歩し、ウェブとアプリ全体で正確に前回の場所から再開できます。
このFastAPI Backend Development Bootcampレッスンでコードを書いて実行できますか?
はい。すべてのFastAPI Backend Development Bootcampレッスンに組み込みコードエディタが含まれているため、ブラウザでリアルコードを書いて実行し、即座のAIフィードバックを取得できます。ローカル設定は不要です。
このコースのすべてのレッスン
- Redisによるキャッシュ戦略
- 非同期データベースアクセス
- ロードバランシングとモニタリング
- バックグラウンドタスクとジョブキュー