Async database with databases library in FastAPI - Performance & Optimization

Performance: Async database with databases library

HIGH IMPACT

This affects how quickly the server can handle database queries without blocking other requests, improving responsiveness and throughput.

Handling database queries in a FastAPI app

FastAPI

import databases
import sqlalchemy
from fastapi import FastAPI

db = databases.Database('sqlite:///test.db')

app = FastAPI()

@app.on_event('startup')
async def startup():
    await db.connect()

@app.on_event('shutdown')
async def shutdown():
    await db.disconnect()

@app.get('/items')
async def read_items():
    query = 'SELECT * FROM items'
    results = await db.fetch_all(query)
    return results

Async calls free the event loop during database operations, allowing other requests to be processed concurrently.

📈 Performance GainNon-blocking queries improve throughput and reduce request latency

Handling database queries in a FastAPI app

FastAPI

import databases
import sqlalchemy
from fastapi import FastAPI

db = databases.Database('sqlite:///test.db')

app = FastAPI()

@app.get('/items')
def read_items():
    query = 'SELECT * FROM items'
    results = db.fetch_all(query)
    return results

Using synchronous database calls blocks the async event loop, causing delays in handling other requests.

📉 Performance CostBlocks event loop during query, increasing response time and reducing concurrency

Performance Comparison

Pattern	DOM Operations	Reflows	Paint Cost	Verdict
Synchronous DB calls in FastAPI	N/A	N/A	N/A	[X] Bad
Async DB calls with databases library	N/A	N/A	N/A	[OK] Good

Rendering Pipeline

Async database calls do not directly affect browser rendering but improve server response time, which reduces time to first byte and interaction delays.

→Server Processing

→Network Response

⚠️ BottleneckServer Processing when synchronous blocking occurs

Core Web Vital Affected

INP

This affects how quickly the server can handle database queries without blocking other requests, improving responsiveness and throughput.

Optimization Tips

1Always use async database calls in async frameworks to avoid blocking.

2Connect and disconnect the database asynchronously during app startup and shutdown.

3Avoid mixing synchronous DB calls inside async endpoints to keep event loop free.

Performance Quiz - 3 Questions

Why is using async database calls important in a FastAPI app?

A It automatically caches all queries.

B It reduces the size of the database.

C It prevents blocking the event loop, improving concurrency and responsiveness.

D It makes the database run faster internally.

DevTools: Network

How to check: Open DevTools, go to Network tab, make a request to the API endpoint, and observe the response time.

What to look for: Look for lower response times and no long blocking periods indicating server-side delays.

Practice

1. What is the main benefit of using the databases library with FastAPI for database operations?

easy

A. It automatically creates database tables without any code.

B. It allows running database queries asynchronously, keeping the app responsive.

C. It replaces FastAPI's routing system with database queries.

D. It makes the database run faster by caching all queries.

Solution

1. Step 1: Understand asynchronous behavior in FastAPI

   FastAPI supports async functions to avoid blocking operations, improving responsiveness.

2. Step 2: Role of the databases library

   The databases library allows async database queries, so the app doesn't wait for the database to respond before continuing.

3. Final Answer:

   It allows running database queries asynchronously, keeping the app responsive. -> Option B

4. Quick Check:

   Async queries keep app responsive = A [OK]

Hint: Async means non-blocking queries for smooth app flow [OK]

Common Mistakes:

• Thinking it auto-creates tables
• Confusing async with caching
• Believing it changes routing

2. Which of the following is the correct way to connect to the database using the databases library in FastAPI?

easy

A. await database.connection()

B. database.connect()

C. database.await_connect()

D. await database.connect()

Solution

1. Step 1: Recall async connection method

   The databases library requires awaiting the connect method because it is asynchronous.

2. Step 2: Identify correct syntax

   The correct syntax is await database.connect(). Calling without await or wrong method names causes errors.

3. Final Answer:

   await database.connect() -> Option D

4. Quick Check:

   Async connect uses await = A [OK]

Hint: Always await async connect calls in databases library [OK]

Common Mistakes:

• Forgetting to use await
• Using wrong method names
• Calling connect synchronously

3. Given this FastAPI code snippet using the databases library, what will be printed when the endpoint is called?

import databases
from fastapi import FastAPI

database = databases.Database('sqlite:///test.db')
app = FastAPI()

@app.on_event('startup')
async def startup():
    await database.connect()

@app.on_event('shutdown')
async def shutdown():
    await database.disconnect()

@app.get('/')
async def read_data():
    query = 'SELECT 1 as number'
    result = await database.fetch_one(query)
    print(result['number'])
    return {'number': result['number']}

medium

A. Error: invalid query

B. None

C. 1

D. Error: database not connected

Solution

1. Step 1: Check database connection lifecycle

   The database connects on startup and disconnects on shutdown, so it is connected when the endpoint runs.

2. Step 2: Analyze query and fetch_one result

   The query selects the number 1 as 'number'. The fetch_one returns a dict-like object with key 'number' and value 1.

3. Final Answer:

   1 -> Option C

4. Quick Check:

   Query returns 1 = D [OK]

Hint: Check connection events and query result keys [OK]

Common Mistakes:

• Assuming no connection established
• Expecting None instead of 1
• Misreading query syntax

4. Identify the error in this FastAPI code using the databases library:

import databases
from fastapi import FastAPI

database = databases.Database('sqlite:///test.db')
app = FastAPI()

@app.on_event('startup')
async def startup():
    await database.connect()

@app.on_event('shutdown')
async def shutdown():
    await database.disconnect()

@app.get('/')
async def read_data():
    query = 'SELECT 1 as number'
    result = database.fetch_one(query)
    return {'number': result['number']}

medium

A. Missing await before database.fetch_one causing a runtime error

B. Database URL is incorrect

C. Missing database.connect() call

D. Using synchronous FastAPI endpoint instead of async

Solution

1. Step 1: Check usage of async database method

   The fetch_one method is async and must be awaited to get the result properly.

2. Step 2: Identify missing await

   Code calls database.fetch_one(query) without await, which causes a runtime error because the coroutine is not awaited.

3. Final Answer:

   Missing await before database.fetch_one causing a runtime error -> Option A

4. Quick Check:

   Async calls need await = C [OK]

Hint: Always await async database calls to avoid errors [OK]

Common Mistakes:

• Forgetting await on async calls
• Assuming fetch_one is synchronous
• Ignoring connection lifecycle

5. You want to fetch all users from a database asynchronously using the databases library in FastAPI. Which code snippet correctly fetches all rows and returns them as a list of dictionaries?

database = databases.Database('sqlite:///test.db')

async def get_users():
    query = 'SELECT * FROM users'
    # Which line correctly fetches all rows?
    ???
    return users

hard

A. users = await database.fetch_all(query)

B. users = database.fetch_all(query)

C. users = await database.fetch_one(query)

D. users = database.fetch_one(query)

Solution

1. Step 1: Understand fetch_all vs fetch_one

   fetch_all returns all rows as a list; fetch_one returns a single row.

2. Step 2: Use await with async fetch_all

   Since fetch_all is async, it must be awaited to get the result.

3. Final Answer:

   users = await database.fetch_all(query) -> Option A

4. Quick Check:

   Fetch all rows async with await = B [OK]

Hint: Use await with fetch_all to get all rows asynchronously [OK]

Common Mistakes:

• Using fetch_one to get all rows
• Forgetting await on async calls
• Calling fetch_all synchronously