1-on-1 Video Calling

Interview-Ready Approach

1) Clarify Scope and SLOs

• •Problem statement: Design a peer-to-peer video calling feature with signaling, NAT traversal, call history, and adaptive quality.
• •Design for a peak load target around 75,000 RPS (including burst headroom).
• •Concurrent calls (peak): ~500,000
• •Daily calls: ~50,000,000
• •Signaling latency: < 1 second
• •Call setup time: < 3 seconds
• •Peer-to-peer success rate: ~80%

2) Capacity Planning Method

• •Convert traffic and growth constraints into request rate, storage growth, and concurrency budgets.
• •Keep at least 2-3x safety margin per tier (ingress, compute, storage, async workers).
• •Reserve explicit latency budgets per hop so p95 can be defended in review.

3) Architecture Decisions

• •WebSockets: Use persistent connection gateways and decouple fanout via pub/sub or queues.
• •Databases: Define a clear system-of-record and design read/write paths separately before adding optimizations.
• •Auth: Centralize identity verification and keep authorization checks close to domain resources.
• •Monitoring: Instrument golden signals (latency, traffic, errors, saturation) per tier and per tenant/domain.

4) Reliability and Failure Strategy

• •Track connection churn, backpressure, and session resumption behavior.
• •Use strong write constraints (transactions or conditional writes) and explicit backup/restore strategy.
• •Use short-lived tokens and secure key rotation workflows.
• •Alert on user-impact SLOs, not only infrastructure metrics.

5) Validation Plan

• •Run one peak-load test, one dependency-degradation test, and one failover test.
• •Verify idempotency for all retried writes and async consumers.
• •Track user-facing SLOs first: p95 latency, error rate, and successful throughput.

6) Trade-offs to Call Out in Interviews

• •WebSockets: WebSockets reduce interaction latency but complicate scaling and state management.
• •Databases: SQL gives stronger transactional guarantees; NoSQL often gives better write scaling and flexibility.
• •Auth: Central auth simplifies policy, but makes auth service availability/security critical.
• •Monitoring: Deep observability speeds incident response but raises ingestion and tooling costs.

Practical Notes

• •The signaling server is lightweight - it just relays SDP/ICE messages. Use WebSockets (already used in the messaging app).
• •Deploy STUN servers globally (they're stateless and cheap). TURN servers are bandwidth-heavy - deploy strategically and cap per-session bandwidth.
• •WebRTC handles encryption (SRTP) natively - no additional E2E encryption needed for the media stream.

Why This Solution Works

Request path: The solution keeps ingress, service logic, and stateful dependencies separated so each layer can scale independently.

Reference flow: Web Clients -> Load Balancer -> API Gateway -> API Service -> Auth Service -> Primary NoSQL DB -> Realtime Bus -> Monitoring

Design strengths

• •Monitoring and logs are wired in from day one for rapid incident triage.
• •Security controls are enforced at ingress to protect downstream capacity.

Interview defense

• •This design makes bottlenecks explicit (ingress, core compute, persistence, async workers).
• •It supports progressive scaling without re-architecting the core request path.
• •It keeps correctness-sensitive state changes in durable systems while offloading background work asynchronously.