Design a Content Delivery for PlanetScale

Question

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

Codemia · Accepted Answer

### Functional Requirements
1. **Content Storage**: The system must store and serve static and dynamic content efficiently.
2. **Low Latency**: The system should deliver content with a latency of under 100ms for 95% of requests.
3. **Multi-Region Support**: Content should be accessible globally with minimal latency, supporting multiple regions and automatic failover.
4. **Versioning**: Ability to serve different versions of content based on user requirements or A/B testing.
5. **Cache Invalidation**: Mechanism for cache invalidation to ensure users receive the latest content.

### Non-Functional Requirements
1. **Scalability**: The system must handle millions of requests per second.
2. **Availability**: Aim for 99.99% uptime with effective disaster recovery strategies.
3. **Consistency**: Support eventual consistency for non-critical content while ensuring strong consistency for critical updates.
4. **Monitoring**: Implement comprehensive monitoring, logging, and alerting mechanisms to detect issues in real-time.

To estimate capacity, we consider:
- **User Base**: Assume 10 million users.
- **Request Frequency**: Each user generates 10 requests per day, leading to 100 million requests daily or approximately 1,157 requests per second.
- **Content Size**: Assume an average content size of 100 KB.
- **Bandwidth**: Total data transfer would be around 11.57 TB/day.

We should provision for peaks, estimating a 10x spike during high traffic events, leading to around 11,570 requests per second during peak times. This demands a robust architecture capable of scaling horizontally to manage load efficiently.

### Components
1. **Load Balancer**: Distributes incoming requests to multiple application servers.
2. **Application Servers**: Handle business logic and content retrieval, deployed in multiple regions (e.g., AWS, GCP).
3. **Content Delivery Network (CDN)**: Caches static content globally to reduce latency.
4. **Database**: PlanetScale Database for dynamic content with horizontal scalability and multi-region capabilities.
5. **Cache Layer**: Redis for in-memory caching of frequently accessed content.

### Technology Choices
- **Load Balancer**: NGINX or AWS ELB
- **Application Servers**: Node.js or Go for high concurrency
- **CDN**: Cloudflare for content distribution
- **Database**: PlanetScale's MySQL-compatible database for OLTP workloads
- **Cache**: Redis for fast access to cached content.

### Schema Design
1. **Content Table**: Stores content metadata and versions.
   - `content_id` (PK, INT)
   - `version` (VARCHAR)
   - `data` (BLOB)
   - `created_at` (TIMESTAMP)
   - `updated_at` (TIMESTAMP)

2. **User Table**: Stores user information and preferences.
   - `user_id` (PK, INT)
   - `preferences` (JSON)

### Access Patterns
- **Read Operations**: High read-to-write ratio, necessitating efficient caching strategies.
- **Write Operations**: Content uploads and updates should be versioned, with recent versions prioritized in cache.

### CAP Theorem Considerations
- **Consistency vs. Availability**: Prioritize availability for non-critical content with eventual consistency, while ensuring strong consistency for critical updates (e.g., user-specific content).
- **Latency vs. Consistency**: Use CDN and edge caching to reduce latency, accepting some degree of eventual consistency for cached content.

### Other Trade-Offs
- **Complexity vs. Performance**: Implementing a multi-region architecture introduces complexity in deployment and data synchronization, but is necessary for performance and user experience.
- **Operational Excellence**: Chaos engineering practices should be in place to test system resilience, alongside SLOs focusing on response time and availability metrics. A phased rollout should start with a minimum viable product (MVP) focusing on a single region before expanding to a global scale.

Design a Content Delivery for PlanetScale

PlanetScale

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

Functional Requirements

Capacity Estimation

Submit Your Answer

PlanetScale 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