Design Video Streaming Infrastructure for IoT devices

Question

Design a high-throughput video streaming system that handles millions of requests. Discuss trade-offs in consistency, availability, and performance.

Codemia · Accepted Answer

### Functional Requirements
1. **Video Streaming**: Support real-time video streaming with low latency for IoT devices.
2. **User Authentication**: Implement secure authentication mechanisms for users and devices accessing the video streams.
3. **Scalability**: Handle millions of concurrent video streams and users without degradation in quality.
4. **Multi-Region Support**: Provide global access with low latency by deploying in multiple regions.
5. **Monitoring**: Enable monitoring of video streams and device performance metrics.

### Non-Functional Requirements
1. **Availability**: The system should guarantee 99.99% uptime.
2. **Latency**: End-to-end latency should be less than 200ms for video streams.
3. **Consistency**: The system should ensure eventual consistency for user data across regions.
4. **Security**: Data encryption in transit and at rest with strict access control policies.

To estimate capacity, consider the following assumptions:
- **IoT Devices**: Assume 1 million devices, each streaming video at 1 Mbps.
- **Concurrent Users**: Assume peak concurrency of 10%. Thus, 100,000 concurrent streams.
- **Data Throughput**: At 1 Mbps, throughput = 100,000 streams * 1 Mbps = 100 Gbps.
- **Storage Needs**: If each device records 24 hours of video at 1 Mbps, that’s:
  - 1 Mbps = 0.125 MB/s
  - 24 hours = 86400 seconds
  - Total storage = 1,000,000 devices * 0.125 MB/s * 86400 seconds = 10.8 PB.
Therefore, the system must support 100 Gbps throughput and manage approximately 10.8 PB of video data.

### Component Diagram
- **IoT Devices**: Streams video to the cloud.
- **Load Balancer**: Distributes incoming streams across multiple streaming servers.
- **Streaming Servers**: Handle video encoding/decoding and distribution. Using **WebRTC** for real-time streaming.
- **Media Storage**: Use **AWS S3** for durable storage of video streams.
- **CDN**: Deploy **Cloudflare** for caching and distributing video content globally.
- **Database**: Use **Cassandra** for user metadata and stream information due to its high write throughput and scalability.
- **Monitoring**: Implement **Prometheus** for metrics collection and **Grafana** for visualizing performance.

### Schema Design
- **Users Table**:  
  - `user_id` (UUID, Primary Key)  
  - `username` (String)  
  - `password_hash` (String)  
  - `device_ids` (List of UUIDs)

- **Devices Table**:  
  - `device_id` (UUID, Primary Key)  
  - `user_id` (UUID, Foreign Key)  
  - `stream_url` (String)  
  - `status` (String)

- **Streams Table**:  
  - `stream_id` (UUID, Primary Key)  
  - `device_id` (UUID, Foreign Key)  
  - `start_time` (Timestamp)  
  - `end_time` (Timestamp)  
  - `duration` (Integer)

This schema supports user access patterns, allowing retrieval of user devices and streams quickly.

### Architectural Decisions
1. **Consistency vs. Availability**: Chose eventual consistency for user data to ensure high availability and scalability. This allows the system to handle massive load while ensuring data is eventually consistent across regions.
2. **Caching Strategy**: Implemented a tiered caching strategy using **Redis** for frequently accessed streams and **CDN** for cache at the edge. This reduces latency significantly but introduces cache invalidation challenges.
3. **Security**: Used OAuth 2.0 for device and user authentication to secure stream access. While this adds overhead, the trade-off is a more secure system against unauthorized access.
4. **Deployment Strategy**: Adopted **blue-green deployment** for seamless updates without downtime, allowing for testing new features in production without impacting existing users.

Design Video Streaming Infrastructure for IoT devices

Twilio

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

Twilio 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