Rate limiting in Ruby on Rails - Performance & Optimization

Performance: Rate limiting

HIGH IMPACT

Rate limiting controls how often users can make requests, affecting server load and response times.

Preventing excessive API requests from users

Ruby on Rails

class ApplicationController < ActionController::Base
  before_action :check_rate_limit

  def check_rate_limit
    if Redis.current.incr(user_rate_key) > 1000
      render plain: 'Rate limit exceeded', status: 429
    end
  end
end

Using Redis for atomic counters reduces CPU load and avoids locking, keeping response times low.

📈 Performance Gainnon-blocking request checks, reduces CPU usage by 70%

Preventing excessive API requests from users

Ruby on Rails

class ApplicationController < ActionController::Base
  before_action :check_rate_limit

  def check_rate_limit
    if request_count_for_user > 1000
      render plain: 'Rate limit exceeded', status: 429
    end
  end
end

Counting requests in memory without efficient storage causes high CPU and memory use, slowing response times.

📉 Performance Costblocks rendering for 50-100ms under high load, increases server CPU usage

Performance Comparison

Pattern	DOM Operations	Reflows	Paint Cost	Verdict
In-memory request counting	0	0	0	[X] Bad
Redis atomic counters for rate limiting	0	0	0	[OK] Good

Rendering Pipeline

Rate limiting happens before rendering; it affects server response time and thus the time to first byte.

→Server Processing

→Network Response

⚠️ BottleneckServer Processing when inefficient rate checks cause delays

Core Web Vital Affected

INP

Rate limiting controls how often users can make requests, affecting server load and response times.

Optimization Tips

1Use external fast stores like Redis for rate limiting counters.

2Avoid in-memory counting that blocks server threads under load.

3Monitor 429 responses and server CPU to tune rate limits.

Performance Quiz - 3 Questions

What is the main performance benefit of using Redis for rate limiting in Rails?

A It delays the first byte sent to the browser

B It increases the number of DOM nodes rendered

C It provides atomic counters that reduce server CPU load

D It triggers more reflows in the browser

DevTools: Network

How to check: Open DevTools, go to Network tab, make rapid requests and observe response status codes and timing.

What to look for: Look for 429 status codes and fast response times indicating efficient rate limiting.

Practice

1. What is the main purpose of rate limiting in a Rails application?

easy

A. To control how often users can make requests to the server

B. To speed up database queries automatically

C. To style the user interface dynamically

D. To manage user sessions securely

Solution

1. Step 1: Understand rate limiting concept

   Rate limiting is about controlling the number of requests a user can make in a time period.

2. Step 2: Identify the main purpose in Rails context

   It helps protect the server from overload and abuse by limiting request frequency.

3. Final Answer:

   To control how often users can make requests to the server -> Option A

4. Quick Check:

   Rate limiting = controlling request frequency [OK]

Hint: Rate limiting controls request frequency to protect servers [OK]

Common Mistakes:

• Confusing rate limiting with UI styling
• Thinking it speeds up database queries
• Mixing it with session management

2. Which of the following is the correct way to configure a simple rate limit using Rack::Attack in Rails?

easy

A. Rack::Attack.throttle('req/ip', limit: 5, period: 1.minute) { |req| req.ip }

B. Rack::Attack.limit('req/ip', 5, 1.minute) { |req| req.ip }

C. Rack::Attack.throttle('req/ip', 1.minute, limit: 5) { |req| req.ip }

D. Rack::Attack.block('req/ip', limit: 5, period: 1.minute) { |req| req.ip }

Solution

1. Step 1: Recall Rack::Attack throttle syntax

   The correct method is throttle with arguments: name, limit, period, and a block returning the key.

2. Step 2: Match syntax with options

   Rack::Attack.throttle('req/ip', limit: 5, period: 1.minute) { |req| req.ip } uses throttle with named parameters limit: 5 and period: 1.minute, and block returning req.ip, which is correct.

3. Final Answer:

   Rack::Attack.throttle('req/ip', limit: 5, period: 1.minute) { |req| req.ip } -> Option A

4. Quick Check:

   Correct method and parameters = Rack::Attack.throttle('req/ip', limit: 5, period: 1.minute) { |req| req.ip } [OK]

Hint: Use throttle with named limit and period parameters [OK]

Common Mistakes:

• Using wrong method name like limit or block
• Swapping limit and period order
• Not passing a block returning the key

3. Given this Rack::Attack throttle configuration:

Rack::Attack.throttle('logins/ip', limit: 3, period: 20.seconds) { |req| req.ip if req.path == '/login' && req.post? }

What happens if a user makes 4 POST requests to '/login' within 20 seconds from the same IP?

medium

A. Only the 1st request is blocked

B. The 4th request is blocked due to rate limit

C. All 4 requests are allowed without restriction

D. The 4th request resets the limit counter

Solution

1. Step 1: Understand throttle parameters

   The limit is 3 requests per 20 seconds per IP for POST requests to '/login'.

2. Step 2: Analyze 4 POST requests from same IP

   The first 3 requests pass, the 4th exceeds the limit and is blocked.

3. Final Answer:

   The 4th request is blocked due to rate limit -> Option B

4. Quick Check:

   Requests > limit = block excess [OK]

Hint: Requests over limit within period get blocked [OK]

Common Mistakes:

• Assuming all requests pass
• Thinking limit resets on 4th request
• Confusing which requests count

4. You wrote this code to limit API requests:

Rack::Attack.throttle('api/ip', limit: 10, period: 1.minute) do |req|
  req.ip if req.path.start_with?('/api')
end

But users report no limits are applied. What is the likely problem?

medium

A. The period should be in seconds, not minutes

B. The limit value 10 is too high to trigger blocking

C. The block returns nil for non-API paths, so no key is generated for those requests

D. The method throttle is not supported in Rack::Attack

Solution

1. Step 1: Check block return value logic

   The block returns req.ip only if path starts with '/api', else returns nil.

2. Step 2: Understand throttle key generation

   If block returns nil, no key is generated and no limit applies for those requests.

3. Final Answer:

   The block returns nil for non-API paths, so no key is generated for those requests -> Option C

4. Quick Check:

   Nil key means no rate limiting applied [OK]

Hint: Block must return key for all requests to apply limits [OK]

Common Mistakes:

• Thinking limit 10 is too high
• Confusing period units
• Believing throttle method is unsupported

5. You want to limit login attempts per user email and IP address separately using Rack::Attack. Which configuration correctly applies this?

hard

A. Rack::Attack.throttle('logins/email', limit: 5, period: 1.hour) { |req| req.params['email'].presence }Rack::Attack.throttle('logins/ip', limit: 10, period: 1.hour) { |req| req.params['email'] }

B. Rack::Attack.throttle('logins/email_ip', limit: 5, period: 1.hour) { |req| req.params['email'] + req.ip }

C. Rack::Attack.throttle('logins/email', limit: 5, period: 1.hour) { |req| req.ip }Rack::Attack.throttle('logins/ip', limit: 10, period: 1.hour) { |req| req.params['email'] }

D. Rack::Attack.throttle('logins/email', limit: 5, period: 1.hour) { |req| req.params['email'] }Rack::Attack.throttle('logins/ip', limit: 10, period: 1.hour) { |req| req.ip }

Solution

1. Step 1: Understand separate throttles for email and IP

   We want one throttle keyed by email, another keyed by IP address.

2. Step 2: Match keys to correct throttle names

   Rack::Attack.throttle('logins/email', limit: 5, period: 1.hour) { |req| req.params['email'] }
   Rack::Attack.throttle('logins/ip', limit: 10, period: 1.hour) { |req| req.ip } correctly uses req.params['email'] for email throttle and req.ip for IP throttle with proper limits.

3. Final Answer:

   Rack::Attack.throttle('logins/email', limit: 5, period: 1.hour) { |req| req.params['email'] }
   Rack::Attack.throttle('logins/ip', limit: 10, period: 1.hour) { |req| req.ip } -> Option D

4. Quick Check:

   Separate keys for email and IP = Rack::Attack.throttle('logins/email', limit: 5, period: 1.hour) { |req| req.params['email'] }
   Rack::Attack.throttle('logins/ip', limit: 10, period: 1.hour) { |req| req.ip } [OK]

Hint: Use separate throttle blocks keyed by email and IP [OK]

Common Mistakes:

• Mixing email and IP keys in wrong throttles
• Concatenating email and IP as one key
• Using IP where email key is needed