CDN concept and usage in HLD - System Design Guide

Problem Statement

When all user requests go directly to a single origin server, it becomes overloaded, causing slow response times and outages. Users far from the server experience high latency, making the service feel sluggish and unreliable.

Solution

A CDN places copies of content on multiple servers distributed worldwide. When a user requests content, the CDN serves it from the nearest server, reducing load on the origin and speeding up delivery by shortening the distance data travels.

Architecture

Origin

Origin

↓

CDN Edge

Server 1

↓

User A

(Near S1)

This diagram shows an origin server distributing content to multiple CDN edge servers. Users receive content from the nearest CDN edge server, reducing latency and origin load.

Trade-offs

✓ Pros

→

Reduces latency by serving content from geographically closer servers.

→

Decreases load on the origin server, improving its stability and scalability.

→

Improves availability by providing multiple points of presence for content delivery.

→

Can reduce bandwidth costs by caching popular content at the edge.

✗ Cons

→

Cache invalidation can be complex, leading to stale content if not managed properly.

→

Initial setup and integration with origin servers add complexity.

→

Less effective for highly dynamic or personalized content that cannot be cached.

Use when serving static or cacheable content to a global or widely distributed user base with request rates above 1,000 requests per second.

Avoid when content is highly dynamic and personalized per user, or when the user base is small and localized, making CDN overhead unnecessary.

Real World Examples

Netflix

Uses CDN to cache and deliver video streams from edge servers close to users, reducing buffering and improving playback quality worldwide.

Amazon

Employs CDN to speed up delivery of product images and static assets globally, ensuring fast page loads for shoppers everywhere.

LinkedIn

Uses CDN to serve static resources like images and scripts, reducing load on core servers and improving user experience across regions.

Alternatives

Origin Pull with Cache-Control Headers

Relies on browsers and intermediate proxies to cache content rather than a distributed network of edge servers.

Use when: When user base is small or regional and deploying a full CDN is not cost-effective.

Peer-to-Peer Content Delivery

Uses users' devices to share content directly, reducing server load without centralized edge servers.

Use when: When distributing large files like software updates where users can share data among themselves.

Summary

CDNs distribute content across multiple edge servers to reduce latency and origin load.

They improve user experience by serving content from locations closer to users worldwide.

CDNs are best for static or cacheable content and large, distributed user bases.

Practice

1. What is the primary purpose of a Content Delivery Network (CDN) in system design?

easy

A. To store copies of content closer to users for faster delivery

B. To increase the size of the main server's database

C. To replace the main server completely

D. To encrypt all user data end-to-end

Solution

1. Step 1: Understand CDN's role in content delivery

   A CDN caches copies of content at locations near users to reduce latency and speed up access.

2. Step 2: Compare options with CDN purpose

   Only To store copies of content closer to users for faster delivery describes storing content near users for faster delivery, which is the main CDN function.

3. Final Answer:

   To store copies of content closer to users for faster delivery -> Option A

4. Quick Check:

   CDN = content near users for speed [OK]

Hint: CDN = cache content near users to speed delivery [OK]

Common Mistakes:

• Thinking CDN replaces main server
• Confusing CDN with encryption service
• Assuming CDN increases database size

2. Which of the following is the correct way to describe how a CDN handles user requests?

easy

A. User requests always go directly to the origin server

B. CDN stores user credentials for authentication

C. CDN modifies user requests to change content

D. User requests are routed to the nearest CDN edge server if content is cached

Solution

1. Step 1: Understand CDN request routing

   CDNs route user requests to the closest edge server that has cached content to reduce latency.

2. Step 2: Evaluate options for request handling

   User requests are routed to the nearest CDN edge server if content is cached correctly states requests go to nearest CDN edge if cached; others are incorrect or unrelated.

3. Final Answer:

   User requests are routed to the nearest CDN edge server if content is cached -> Option D

4. Quick Check:

   CDN routes requests to nearest cache [OK]

Hint: Requests go to nearest cached CDN server, not origin [OK]

Common Mistakes:

• Assuming all requests hit origin server
• Believing CDN changes content on the fly
• Confusing CDN with authentication service

3. Consider a system using a CDN with edge servers in multiple regions. If a user in Europe requests a static image, which server will most likely serve the image?

medium

A. A CDN edge server located in Europe

B. The origin server located in the US

C. A random CDN edge server in Asia

D. The user's local device cache only

Solution

1. Step 1: Identify CDN edge server selection logic

   CDNs serve content from the closest edge server to the user to reduce latency.

2. Step 2: Match user location with server location

   A user in Europe will get content from a European CDN edge server if available, not from origin or distant servers.

3. Final Answer:

   A CDN edge server located in Europe -> Option A

4. Quick Check:

   User location matches nearest CDN edge server [OK]

Hint: User served by nearest CDN edge server geographically [OK]

Common Mistakes:

• Assuming origin server always serves content
• Choosing random or distant CDN servers
• Ignoring CDN edge server caching

4. A developer notices that users are still experiencing slow load times despite using a CDN. Which of the following is the most likely cause?

medium

A. The CDN is encrypting content which slows delivery

B. The CDN is not caching the content properly, causing requests to go to the origin server

C. The CDN is blocking user requests randomly

D. The CDN is storing user passwords insecurely

Solution

1. Step 1: Analyze slow load cause with CDN

   If CDN cache misses occur, requests go to origin server, increasing latency and slowing load times.

2. Step 2: Evaluate options for likely cause

   The CDN is not caching the content properly, causing requests to go to the origin server explains cache miss causing slow loads; others are unrelated or incorrect.

3. Final Answer:

   The CDN is not caching the content properly, causing requests to go to the origin server -> Option B

4. Quick Check:

   Cache miss = slow load [OK]

Hint: Slow CDN? Check if cache is missing content [OK]

Common Mistakes:

• Blaming encryption for speed issues
• Assuming CDN blocks requests randomly
• Confusing CDN with security storage

5. You are designing a global video streaming service. Which CDN usage strategy best improves performance and reduces origin server load?

hard

A. Disable CDN caching to ensure fresh content always

B. Cache only small static files like logos, serve videos directly from origin

C. Cache all static content including videos on CDN edge servers worldwide

D. Use CDN only for user authentication requests

Solution

1. Step 1: Identify content types for CDN caching

   Videos are large static files that benefit greatly from CDN caching near users to reduce latency and origin load.

2. Step 2: Evaluate CDN strategies for global streaming

   Caching all static content including videos on CDN edges worldwide improves performance and reduces origin server load.

3. Final Answer:

   Cache all static content including videos on CDN edge servers worldwide -> Option C

4. Quick Check:

   Cache large static files globally for best performance [OK]

Hint: Cache all static content globally on CDN edges [OK]

Common Mistakes:

• Caching only small files and ignoring videos
• Disabling CDN caching unnecessarily
• Using CDN for unrelated tasks like authentication