What is the best way to practice mock system design interviews?

The best way to practice mock system design interviews is to use a structured platform like System Design Lab. It provides real interview questions (design URL shortener, design Twitter, design Netflix), an interactive diagram builder, an AI interviewer that asks follow-up questions, and detailed feedback on your architecture, scalability decisions, and trade-offs — exactly like a real FAANG interview.

How do I prepare for a system design interview?

To prepare for a system design interview: (1) Learn core concepts like distributed systems, caching, databases, sharding, and message queues. (2) Practice drawing architecture diagrams under time pressure. (3) Take knowledge quizzes to test your understanding. (4) Do mock system design interviews with AI feedback to get comfortable explaining your decisions. (5) Review community solutions to see how others approach the same problems. System Design Lab covers all five steps in one platform.

What system design interview questions should I practice?

Common system design interview questions include: Design a URL shortener, Design Twitter/social media feed, Design Netflix/video streaming, Design a distributed cache, Design a rate limiter, Design a notification system, Design a ride-sharing app like Uber. System Design Lab offers a growing bank of real interview problems — Twitter, YouTube, URL shortener, Uber and more — covering the major categories asked at FAANG and top-tier tech companies.

Is there a free mock system design interview tool?

Yes — System Design Lab offers a free 7-day trial with access to 3 mock system design interview problems, all learning modules, all quizzes, and community solutions. No credit card required. Premium (₹999 for 90 days) unlocks unlimited problems and AI interviewer access.

How does System Design Lab's AI interview feedback work?

After you submit your system design, the AI evaluates your written explanation and architecture diagram together. It scores you on completeness, scalability, fault tolerance, and clarity, then gives you specific, actionable feedback — similar to what a senior engineer would say after your interview. You can also chat with an AI interviewer in real-time during your attempt.

System Design Lab›System Design Questions›Design Live Comments

Design Live Comments

MediumReal-time Systemsreal-timewebsocketsmessagingrate-limiting🟣 Meta

Question Overview

Design live comments for a live-video platform, where every viewer of a broadcast sees new comments within seconds. The hard parts are routing comments only to the right connections, holding tens of millions of persistent connections, and keeping a stream with millions of viewers readable and stable.…

Requirements

Post a comment on a live video and have it appear for other viewers within 2 seconds at p95
Viewers joining mid-stream see the most recent comments at once, then new ones live
Broadcasters and moderators delete comments and ban users, with removals propagated to connected viewers
Per-user rate limits and spam filtering on the write path
Stay stable when one stream has 5M concurrent viewers; sampling comments on busy streams is acceptable
Acknowledged comments are durably stored for replay; the live feed favors availability over strict ordering

Back-of-the-envelope numbers

Comment writes: 10K/s at peak, ~4K/s average × 86,400 s ≈ 350M comments/day × 300 B ≈ 100 GB/day ≈ 38 TB/year
Naive hot-stream fan-out: 2K comments/s × 5M viewers = 10B deliveries/s, impossible to push and unreadable by any human
Sampled hot stream: cap at 10 comments/s per viewer, batched into 2 frames/s of ~1.5 KB → 5M × 2 = 10M frames/s ≈ 15 GB/s (120 Gbps)
Connections: 50M concurrent WebSocket or SSE connections ÷ ~100K per gateway host ≈ 500 hosts, ~1,000 with failover headroom
Pub/sub fan-out: the hot stream's viewers sit on at most ~1,000 gateways, so each comment batch is published ≤ 1,000 times, not 5M times
Join backfill cache: last 50 comments × 1M live streams × 300 B ≈ 15 GB, easily held in a Redis cluster

Key components

Comment service that validates, rate-limits per user, runs spam and profanity checks, persists the comment, then publishes it to the video's channel
Pub/sub layer with one channel per live video, so each comment is routed only to gateways that have viewers of that video
Real-time gateways holding WebSocket or SSE connections, with an in-memory map of video_id to local connections for last-mile fan-out
Batching and sampling on hot streams: gateways flush every ~500 ms and cap comments per viewer, favoring friends, verified accounts, and highlighted comments
Recent-comments cache: a capped list per video in Redis that serves the last 50 comments to viewers joining mid-stream
Moderation events such as deletes and bans flow through the same channels as tombstones so clients remove comments already on screen
Durable comment store (e.g. Cassandra keyed by video_id and time bucket) that also powers replay against the recording's timestamps

Common mistakes

Having clients poll the database for new comments every second, turning 50M viewers into 50M mostly empty queries per second
Pushing every comment to every viewer of a hot stream: 2K comments/s × 5M viewers is 10B messages/s, far beyond what anyone can read
Broadcasting every comment to every gateway instead of routing by video subscription, so each gateway processes all 10K comments/s
Requiring strict global ordering or strong consistency for comments, adding latency to a feature that tolerates approximate order
Letting sampling hide a viewer's own comment; the author should always see it immediately through a local echo
Pinning each video to a single server for all writes and fan-out, so one hot stream overloads one machine

Likely follow-ups

How would you choose which comments to show when a stream receives thousands per second?
How would you replay comments in sync with the recorded video after the broadcast ends?
What happens to viewers when a gateway server crashes, and how do they reconnect without missing comments?
How would you add reactions such as hearts, which arrive at many times the comment rate?
How would you serve viewers of the same stream spread across several regions?
How would you shadow-ban a spammer so they still see their own comments but nobody else does?

No community solutions yet

Be the first to publish your solution

Practice ‘Design Live Comments’ with an AI Interviewer

Get scored feedback on your diagram, scalability approach, and trade-offs. Free while we grow — up to 3 full interviews a day.

Try a 5-minute drill — no signup