G object for request-scoped data in Flask - Performance & Optimization

Performance: G object for request-scoped data

MEDIUM IMPACT

This affects how data is stored and accessed during a single web request, impacting memory usage and request handling speed.

Storing and accessing data during a single request lifecycle

Flask

from flask import g

def get_user():
    if not hasattr(g, 'user'):
        g.user = compute_user_from_db()
    return g.user

@app.route('/')
def index():
    user1 = get_user()
    user2 = get_user()
    return f"Hello {user1.name} and {user2.name}"

Caches the user data in the request-scoped G object, avoiding repeated database calls.

📈 Performance GainSaves 50-100ms by reducing duplicate DB queries per request

Storing and accessing data during a single request lifecycle

Flask

def get_user():
    user = compute_user_from_db()
    return user

@app.route('/')
def index():
    user1 = get_user()
    user2 = get_user()
    return f"Hello {user1.name} and {user2.name}"

Calling the same expensive function multiple times per request causes repeated database queries and slows response.

📉 Performance CostBlocks rendering for extra 50-100ms per repeated call depending on DB latency

Performance Comparison

Pattern	Memory Usage	Request Time	Server Load	Verdict
Repeated function calls without G	Low per call but repeated	High due to duplicate DB queries	Higher due to redundant work	[X] Bad
Caching data in G object	Slightly higher per request	Lower by avoiding repeats	Lower due to efficient reuse	[OK] Good

Rendering Pipeline

The G object stores data in memory during the request lifecycle, so it mainly affects server-side request processing before the response is sent.

→Request Handling

→Server Processing

⚠️ BottleneckRepeated expensive computations or database calls without caching

Core Web Vital Affected

INP

This affects how data is stored and accessed during a single web request, impacting memory usage and request handling speed.

Optimization Tips

1Cache expensive computations or database calls in the G object during a request.

2Avoid calling the same function multiple times without caching in G.

3Use G only for data valid during a single request to prevent memory leaks.

Performance Quiz - 3 Questions

What is the main performance benefit of using Flask's G object for request-scoped data?

A Avoids repeating expensive computations during the same request

B Reduces the size of the final HTML sent to the client

C Improves browser rendering speed by caching CSS

D Allows sharing data between different user sessions

DevTools: Network and Performance panels

How to check: Record a request in Performance panel and look for repeated database or function call durations; check Network panel for request timing.

What to look for: Look for duplicated long-running calls or delays in server response time indicating redundant processing.

Practice

1. What is the main purpose of the g object in Flask?

easy

A. To store data temporarily during a single request

B. To store data permanently across all requests

C. To manage database connections globally

D. To handle user sessions across multiple requests

Solution

1. Step 1: Understand the scope of g

   The g object is designed to hold data only during the life of one request.

2. Step 2: Compare with other storage options

   Unlike session or global variables, g does not persist data beyond the current request.

3. Final Answer:

   To store data temporarily during a single request -> Option A

4. Quick Check:

   g = request-scoped storage [OK]

Hint: Remember: g lasts only for one request [OK]

Common Mistakes:

• Thinking g stores data across requests
• Confusing g with session or global variables
• Assuming g manages database connections automatically

2. Which of the following is the correct way to store a value in the g object inside a Flask view function?

easy

A. g.user = 'Alice'

B. g.add('user', 'Alice')

C. g.set('user', 'Alice')

D. g['user'] = Alice

Solution

1. Step 1: Recall how to assign attributes in g

   g acts like a simple object where you assign attributes using dot notation.

2. Step 2: Check each option's syntax

   g.user = 'Alice' uses g.user = 'Alice', which is the correct way. g['user'] = Alice uses dictionary style which is invalid. Options B and C use methods that g does not have.

3. Final Answer:

   g.user = 'Alice' -> Option A

4. Quick Check:

   Use dot notation to set g attributes [OK]

Hint: Use dot notation: g.attribute = value [OK]

Common Mistakes:

• Using dictionary syntax like g['key']
• Trying to call non-existent methods on g
• Confusing g with Flask's session object

3. Consider this Flask code snippet:

from flask import Flask, g
app = Flask(__name__)

@app.before_request
def before():
    g.value = 10

@app.route('/')
def index():
    g.value += 5
    return str(g.value)

What will be the output when you visit the '/' route?

medium

A. 10

B. 5

C. 15

D. Error: AttributeError

Solution

1. Step 1: Understand the before_request function

   Before each request, g.value is set to 10.

2. Step 2: Analyze the index function

   It adds 5 to g.value, making it 15, then returns it as a string.

3. Final Answer:

   15 -> Option C

4. Quick Check:

   g.value starts 10, adds 5 = 15 [OK]

Hint: Remember g resets each request; additions apply after setup [OK]

Common Mistakes:

• Assuming g.value keeps old values between requests
• Forgetting before_request runs before route
• Expecting an error due to missing attribute

4. What is wrong with this Flask code using g?

from flask import Flask, g
app = Flask(__name__)

@app.route('/')
def index():
    if not hasattr(g, 'counter'):
        g.counter = 0
    g.counter += 1
    return str(g.counter)

medium

A. It will raise an error because g cannot have new attributes

B. The code will cause a syntax error due to missing import

C. The counter will increase across requests, causing unexpected results

D. The counter resets on every request, so it always returns 1

Solution

1. Step 1: Understand g scope

   g is reset for each request, so counter does not persist between requests.

2. Step 2: Analyze the counter behavior

   Each request starts fresh, so counter is set to 0 then incremented to 1 every time.

3. Final Answer:

   The counter resets on every request, so it always returns 1 -> Option D

4. Quick Check:

   g data is request-scoped, no persistence [OK]

Hint: Remember g resets each request; no data persists [OK]

Common Mistakes:

• Expecting g.counter to keep increasing across requests
• Thinking g forbids new attributes
• Confusing g with session or global variables

5. You want to share a database connection during a single Flask request using g. Which code snippet correctly opens the connection once per request and reuses it in multiple functions?

from flask import Flask, g
app = Flask(__name__)

def get_db():
    if not hasattr(g, 'db'):
        g.db = open_database_connection()
    return g.db

@app.route('/')
def index():
    db = get_db()
    # use db
    return 'OK'

What is the main benefit of this pattern?

hard

A. It opens a new database connection on every function call

B. It shares one connection per request, avoiding multiple opens

C. It stores the connection globally for all requests

D. It closes the connection automatically after each function

Solution

1. Step 1: Analyze the get_db function

   It checks if g.db exists; if not, it opens the connection and saves it to g.

2. Step 2: Understand request scope of g

   The connection is stored only for the current request, so multiple calls reuse the same connection.

3. Step 3: Compare options

   It shares one connection per request, avoiding multiple opens correctly describes this behavior. It opens a new database connection on every function call is wrong because it avoids multiple opens. It stores the connection globally for all requests is wrong because g is not global. It closes the connection automatically after each function is incorrect; closing is not automatic here.

4. Final Answer:

   It shares one connection per request, avoiding multiple opens -> Option B

5. Quick Check:

   Use g to cache per-request resources [OK]

Hint: Cache resources in g to reuse during request [OK]

Common Mistakes:

• Thinking g stores data globally across requests
• Opening new connections on every function call
• Assuming g auto-closes resources