System Design: Designing a Video Conferencing System (Zoom / MS Teams)

System Design: Designing a Video Conferencing System

Mental Model

Connecting isolated components into a resilient, scalable, and observable distributed web.

Designing a real-time video conferencing system like Zoom or Microsoft Teams is fundamentally different from a video streaming service like YouTube. While YouTube prioritizes quality and high resolution, Zoom prioritizes Latency. A delay of more than 150ms makes a conversation impossible.

1. Core Requirements

graph LR
    Producer[Producer Service] -->|Publish Event| Kafka[Kafka / Event Bus]
    Kafka -->|Consume| Consumer1[Consumer Group A]
    Kafka -->|Consume| Consumer2[Consumer Group B]
    Consumer1 --> DB1[(Primary DB)]
    Consumer2 --> Cache[(Redis)]

Real-time Video/Audio: Bi-directional streaming with sub-200ms latency.
Large Meetings: Supporting hundreds or thousands of participants.
Screen Sharing: Sharing a high-resolution, low-framerate stream.
Resilience: Handling varying network conditions (packet loss, low bandwidth).

2. The Protocol: UDP vs. TCP

The Choice:UDP (User Datagram Protocol) is the mandatory choice for real-time media.
Why? TCP's error correction (re-sending lost packets) causes delay. In a call, it's better to lose a single frame of video (a minor glitch) than to pause the whole call to wait for that frame to arrive.

3. Communication Technology: WebRTC

WebRTC is the standard for real-time communication in the browser. It handles:

STUN/TURN Servers: For bypassing firewalls and finding the best path between peers.
Signaling: Using WebSockets to exchange metadata (like "I'm calling you") before the media starts flowing.

4. Scaling the Meeting: SFU vs. MCU

How do you deliver 100 video streams to 100 participants?

Option A: Peer-to-Peer (Mesh)

Every user sends their stream to every other user.

Limit: Only works for 2-3 people. A user's upload bandwidth will crash with more.

Option B: MCU (Multipoint Control Unit)

The server receives all streams, mixes them into one single video (like a collage), and sends that one stream to everyone.

Pros: Low bandwidth for the client.
Cons: Extremely CPU-intensive for the server.

Option C: SFU (Selective Forwarding Unit) - The Standard

The server receives all streams but doesn't mix them. It simply forwards the relevant streams to each participant.

The Optimization: If a participant is muted and their camera is off, the SFU stops forwarding their data. This is how Zoom scales to 1,000 people.

5. Handling Network Jitter (Adaptive Bitrate)

Simulcast: The client sends three versions of their video (High, Medium, Low quality) to the SFU. The SFU forwards the High-quality version to users with fast internet and the Low-quality version to users with slow mobile data.

6. Global Scalability

Video servers must be placed in data centers geographically close to participants to minimize the "Speed of Light" delay.

Geo-routing: If users in London are talking, the meeting should be hosted on a server in London, not New York.

Summary

The engineering of video conferencing is a masterclass in Low-latency Networking. By leveraging UDP, SFU architectures, and Simulcast for adaptive quality, you can build a platform that makes global communication feel as natural as a face-to-face meeting.

Engineering Standard: The "Staff" Perspective

In high-throughput distributed systems, the code we write is often the easiest part. The difficulty lies in how that code interacts with other components in the stack.

1. Data Integrity and The "P" in CAP

Whenever you are dealing with state (Databases, Caches, or In-memory stores), you must account for Network Partitions. In a standard Java microservice, we often choose Availability (AP) by using Eventual Consistency patterns. However, for financial ledgers, we must enforce Strong Consistency (CP), which usually involves distributed locks (Redis Redlock or Zookeeper) or a strictly linearizable sequence.

2. The Observability Pillar

Writing logic without observability is like flying a plane without a dashboard. Every production service must implement:

Tracing (OpenTelemetry): Track a single request across 50 microservices.
Metrics (Prometheus): Monitor Heap usage, Thread saturation, and P99 latencies.
Structured Logging (ELK/Splunk): Never log raw strings; use JSON so you can query logs like a database.

3. Production Incident Prevention

To survive a 3:00 AM incident, we use:

Circuit Breakers: Stop the bleeding if a downstream service is down.
Bulkheads: Isolate thread pools so one failing endpoint doesn't crash the entire app.
Retries with Exponential Backoff: Avoid the "Thundering Herd" problem when a service comes back online.

Critical Interview Nuance

When an interviewer asks you about this topic, don't just explain the code. Explain the Trade-offs. A Staff Engineer is someone who knows that every architectural decision is a choice between two "bad" outcomes. You are picking the one that aligns with the business goal.

Performance Checklist for High-Load Systems:

Minimize Object Creation: Use primitive arrays and reusable buffers.
Batching: Group 1,000 small writes into 1 large batch to save I/O cycles.
Async Processing: If the user doesn't need the result immediately, move it to a Message Queue (Kafka/SQS).

Advanced Architectural Blueprint: The Staff Perspective

In modern high-scale engineering, the primary differentiator between a Senior and a Staff Engineer is the ability to see beyond the local code and understand the Global System Impact. This section provides the exhaustive architectural context required to operate this component at a "MANG" (Meta, Amazon, Netflix, Google) scale.

1. High-Availability and Disaster Recovery (DR)

Every component in a production system must be designed for failure. If this component resides in a single availability zone, it is a liability.

Multi-Region Active-Active: To achieve "Five Nines" (99.999%) availability, we replicate state across geographical regions using asynchronous replication or global consensus (Paxos/Raft).
Chaos Engineering: We regularly inject "latency spikes" and "node kills" using tools like Chaos Mesh to ensure the system gracefully degrades without a total outage.

2. The Data Integrity Pillar (Consistency Models)

When managing state, we must choose our position on the CAP theorem spectrum.

Model	latency	Complexity	Use Case
Strong Consistency	High	High	Financial Ledgers, Inventory Management
Eventual Consistency	Low	Medium	Social Media Feeds, Like Counts
Monotonic Reads	Medium	Medium	User Profile Updates

3. Observability and "Day 2" Operations

Writing the code is only 10% of the lifecycle. The remaining 90% is spent monitoring and maintaining it.

Tracing (OpenTelemetry): We use distributed tracing to map the request flow. This is critical when a P99 latency spike occurs in a mesh of 100+ microservices.
Structured Logging: We avoid unstructured text. Every log line is a JSON object containing correlationId, tenantId, and latencyMs.
Custom Metrics: We export business-level metrics (e.g., "Orders processed per second") to Prometheus to set up intelligent alerting with PagerDuty.

4. Production Readiness Checklist for Staff Engineers

Capacity Planning: Have we performed load testing to find the "Breaking Point" of the service?
Security Hardening: Is all communication encrypted using mTLS (Mutual TLS)?
Backpressure Propagation: Does the service correctly return HTTP 429 or 503 when its internal thread pools are saturated?
Idempotency: Can the same request be retried 10 times without side effects? (Critical for Payment systems).

Critical Interview Reflection

When an interviewer asks "How would you improve this?", they are looking for your ability to identify Bottlenecks. Focus on the network I/O, the database locking strategy, or the memory allocation patterns of the JVM. Explain the trade-offs between "Throughput" and "Latency." A Staff Engineer knows that you can never have both at their theoretical maximums.

Optimization Summary:

Reduce Context Switching: Use non-blocking I/O (Netty/Project Loom).
Minimize GC Pressure: Prefer primitive specialized collections over standard Generics.
Data Sharding: Use Consistent Hashing to avoid "Hot Shards."

Technical Trade-offs: Messaging Systems

Pattern	Ordering	Durability	Throughput	Complexity
Log-based (Kafka)	Strict (per partition)	High	Very High	High
Memory-based (Redis Pub/Sub)	None	Low	High	Very Low
Push-based (RabbitMQ)	Fair	Medium	Medium	Medium

Key Takeaways

Real-time Video/Audio: Bi-directional streaming with sub-200ms latency.
Large Meetings: Supporting hundreds or thousands of participants.
Screen Sharing: Sharing a high-resolution, low-framerate stream.

Verbal Interview Script

Interviewer: "How would you ensure high availability and fault tolerance for this specific architecture?"

Candidate: "To achieve 'Five Nines' (99.999%) availability, we must eliminate all Single Points of Failure (SPOF). I would deploy the API Gateway and stateless microservices across multiple Availability Zones (AZs) behind an active-active load balancer. For the data layer, I would use asynchronous replication to a read-replica in a different region for disaster recovery. Furthermore, it's not enough to just deploy redundantly; we must protect the system from cascading failures. I would implement strict timeouts, retry mechanisms with exponential backoff and jitter, and Circuit Breakers (using a library like Resilience4j) on all synchronous network calls between microservices."

System Design: Designing a Video Conferencing System (Zoom / MS Teams)