Design a scalable Content Delivery System

Question

Design a scalable content delivery system that handles millions of requests. Discuss trade-offs in consistency, availability, and performance.

Codemia · Accepted Answer

- **Functional Requirements**:
  - Support for delivering various types of content (e.g., images, videos, documents) to users with low latency.
  - RESTful API to retrieve content, with support for pagination and filtering.
  - Rate limiting to prevent abuse and ensure fair usage among users.
  - Caching mechanisms to reduce load on backend services and improve response times.

- **Non-Functional Requirements**:
  - **Scalability**: Must handle millions of requests per day, scaling horizontally as needed.
  - **Availability**: Aim for 99.99% uptime with a system designed to handle failures gracefully.
  - **Performance**: Response time should be under 200ms for 95% of requests.
  - **Consistency**: Use eventual consistency for cache updates to allow for higher availability.

Assuming Coinbase has approximately 5 million active users, and each user makes an average of 10 requests per day:
- **Total Requests per Day**: 5 million users * 10 requests = 50 million requests/day.
- **Average Request Rate**: 50 million requests/day / 86400 seconds = ~578 requests/second.
- Considering peaks, we can estimate a peak load of 3x average, leading to ~1734 requests/second during peak times.
- **Storage Needs**: If we assume an average content size of 1MB, and a caching strategy to hold 80% of requests, we would need around 1.5TB of cache storage.

The architecture consists of several key components:
- **Load Balancer**: Distributes incoming requests to multiple web servers.
- **Web Servers**: Handle API requests, performing authentication and authorization.
- **Content Delivery Network (CDN)**: For static content, reducing latency by caching content closer to users.
- **Cache Layer**: Using Redis or Memcached to store frequently accessed content for quick retrieval.
- **Database**: A distributed NoSQL database like Cassandra for storing metadata and user content, ensuring scalability and availability.
- **Monitoring & Logging**: Tools like Prometheus for monitoring and ELK stack for logging to ensure observability and alerting.

The database schema could look like this:
- **Users Table**: `user_id (PK)`, `username`, `email`, `created_at`, `updated_at`
- **Content Table**: `content_id (PK)`, `user_id (FK)`, `content_type`, `content_url`, `metadata`, `created_at`
- **Access Logs Table**: `log_id (PK)`, `user_id (FK)`, `content_id (FK)`, `timestamp`, `response_time`

**Access Patterns**:
- Frequent reads from Users and Content tables through indexed queries.
- Logging access patterns for analytics and potential caching strategies.

- **Consistency vs. Availability**: We prioritize availability with eventual consistency in caching, allowing for higher uptime, but slight delays in data updates.
- **Cost vs. Performance**: While using a CDN and cache improves performance, it may increase costs. We balance by caching only the most accessed content.
- **Complexity vs. Scalability**: A distributed database like Cassandra offers scalability but adds complexity in terms of data modeling and consistency management. Simplicity could be achieved with a relational database, but that would limit scaling capabilities.
- **Monitoring vs. Performance Overhead**: Comprehensive monitoring tools add overhead, but are critical for proactively catching issues. We must decide on the granularity of metrics to monitor, balancing detail with performance.

Design a scalable Content Delivery System

Coinbase

What the Interviewer Expects

Key Topics to Cover

How to Approach This

Possible Follow-up Questions

Practice a Similar Problem on Codemia

Sample Answer

Requirements

Capacity Estimation

Submit Your Answer

Coinbase Software Engineer Interview Guide

Related Questions

Design a high-throughput Inventory Management System

Design a low-latency Rate Limiting System

Design a fault-tolerant Payment System

Design a fault-tolerant Messaging System

Design Walmart Product Search