teacher
Sign in
Track

System Design

How real systems stay fast, available and correct — explained from first principles.

46 lessons~25 hours of lessonsTarget: Intermediate — confident in a 45-minute system design interview
Reality check
  • Articles alone will not make you good at system design. Build small systems, break them, and read real postmortems.
  • Interviews test reasoning and trade-offs, not memorised diagrams. Two candidates with the same diagram can get opposite results.
  • Numbers matter. If you cannot estimate QPS, storage and bandwidth roughly, your design is a guess.
  • Most real-world design is choosing what not to build. Simple and boring usually wins.
  1. Stage 10/6

    Foundations

    How a request travels — networks, DNS, HTTP, APIs — and how to estimate load.

    1. 01Start HereWhat system design is, how this track is laid out, what interviews actually test, and how to practise without fooling yourself.15 min
    2. 02Networking BasicsIP addresses, ports, TCP versus UDP, the difference between latency and bandwidth, and the latency numbers worth remembering.25 min
    3. 03DNS and HTTPHow a name becomes an IP address, what an HTTP request contains, what TLS adds, and how HTTP/1.1, 2 and 3 differ.30 min
    4. 04APIsREST, gRPC and GraphQL for request-response, and WebSockets, long polling and SSE for pushing data to clients.30 min
    5. 05Back-of-the-Envelope EstimationPowers of two, the one-day-is-1e5-seconds rule, and quick QPS, storage and bandwidth estimates that guide a design.30 min
    6. 06The Interview FrameworkA repeatable 45-minute plan for a system design interview, with the mistakes that cost candidates the most.25 min
  2. Stage 20/10

    Building blocks

    Scaling, load balancing, caching, CDNs and every major database idea.

    1. 01Scaling BasicsVertical versus horizontal scaling, why stateless services scale easily, and where the state has to go instead.20 min
    2. 02Load BalancingHow a load balancer spreads traffic across servers, the L4 versus L7 choice, common algorithms, health checks and sticky sessions.25 min
    3. 03CachingKeep a copy of hot data in fast memory so most requests never reach the database.30 min
    4. 04CDNs and ProxiesPut servers between users and your origin - to hide, filter and cache - and push static content close to the user.25 min
    5. 05Databases - SQL vs NoSQLWhat ACID and isolation levels guarantee, the main NoSQL families, and a practical way to choose.30 min
    6. 06IndexingHow B+tree, hash, composite and covering indexes speed up reads, what they cost on writes, and where LSM trees and inverted indexes fit.30 min
    7. 07ReplicationKeep copies of data on several machines - leader-follower, sync versus async, lag, multi-leader and quorums.30 min
    8. 08Sharding and PartitioningSplit a dataset across machines - range versus hash, choosing a shard key, hot keys and resharding without downtime.30 min
    9. 09Consistent HashingPlace keys and servers on a ring so adding or removing a server moves only a small share of the keys.25 min
    10. 10CAP and ConsistencyWhat the CAP theorem really says, how PACELC adds latency, and the consistency levels from strong to eventual.30 min
  3. Stage 30/6

    Distributed patterns

    Queues, rate limiting, IDs, reliability, microservices, observability and security.

    1. 01Message Queues and Pub/SubPut a durable buffer between services so they can work at different speeds, and learn how Kafka partitions, offsets and delivery guarantees behave.35 min
    2. 02Rate LimitingCap how many requests a client can make, using token bucket, leaky bucket and window counters, usually backed by Redis.30 min
    3. 03Unique ID GenerationHow distributed systems create IDs that never collide, from auto-increment and UUIDs to time-ordered Snowflake IDs.25 min
    4. 04Reliability and AvailabilityMeasure uptime with SLIs, SLOs and nines, then survive failures with retries, backoff, idempotency keys and circuit breakers.35 min
    5. 05Microservices and API GatewayWhen splitting a system into services pays off, how services find each other, what a gateway does, and how sagas replace distributed transactions.30 min
    6. 06Observability and SecuritySee what a running system is doing with logs, metrics and traces, and secure it with authentication, OAuth 2.0, JWTs, TLS and secret management.35 min
  4. Stage 40/8

    Real-world tools

    How Kafka, Redis, Elasticsearch, Cassandra, Nginx, S3, ZooKeeper and Kubernetes actually work inside, and when to pick each.

    1. 01Kafka InternalsHow Kafka stores messages in replicated partition logs, and how producers, consumer groups, offsets and transactions behave under failure.45 min
    2. 02Redis InternalsWhy a single-threaded in-memory store is so fast, which data structures it offers, and how persistence, replication and clustering work.40 min
    3. 03Elasticsearch InternalsHow an inverted index, shards and immutable segments make full-text search and log analytics fast, and why it should not be your primary database.40 min
    4. 04Cassandra and DynamoDBHow wide-column stores spread data over a ring, trade consistency for availability, write through LSM trees, and why you model data from the queries backwards.45 min
    5. 05Nginx and Reverse ProxiesHow Nginx handles thousands of connections with a few event-driven workers, and how to use it for load balancing, TLS, caching and rate limiting.35 min
    6. 06Object Storage and Amazon S3How S3 stores files as objects under keys, what its durability and consistency promises really mean, and how to control its costs.35 min
    7. 07ZooKeeper and etcdWhy distributed systems need a small, strongly consistent coordination service, how Raft and ZAB keep it correct, and how leader election, locks and discovery use it.35 min
    8. 08Docker and KubernetesHow containers package software, how Kubernetes schedules and heals many of them, and an honest view of when you do not need it.40 min
  5. Stage 50/16

    Case studies

    End-to-end designs of systems you use every day.

    1. 01Design a URL ShortenerA full case study of a TinyURL-style service, covering short-code generation, a read-heavy store, caching and redirect trade-offs.35 min
    2. 02Design a Chat SystemA full case study of a WhatsApp-style messenger, covering WebSocket connections, message ordering, delivery guarantees, offline users and presence.40 min
    3. 03Design a News FeedA full case study of a social feed, covering fan-out on write versus read, the celebrity problem, feed caches and pagination.35 min
    4. 04Design a Notification SystemA full case study of a push, SMS and email notification platform, covering queues, retries, deduplication, priorities and user preferences.30 min
    5. 05Design a Video Streaming ServiceA full case study of a YouTube-style platform, covering upload, transcoding pipelines, adaptive bitrate streaming and CDN delivery.35 min
    6. 06Design a Ride-Sharing ServiceA full case study of an Uber-style service, covering high-rate location updates, geospatial indexes, driver matching and the trip state machine.35 min
    7. 07Design a Payment SystemA full case study of a payment backend, covering idempotency, double-entry ledgers, unknown outcomes from banks and daily reconciliation.40 min
    8. 08Design Search AutocompleteA full case study of typeahead suggestions, covering tries with cached top-k lists, offline index builds, caching and sharding by prefix.30 min
    9. 09Design a Web CrawlerA full case study of a crawler, covering the URL frontier, politeness, robots.txt, duplicate detection with Bloom filters, DNS caching and recrawl scheduling.35 min
    10. 10Design a Distributed Key-Value StoreA Dynamo-style design covering partitioning with consistent hashing, replication, quorum reads and writes, vector clocks, hinted handoff, Merkle-tree repair and gossip.40 min
    11. 11Design File Storage and SyncA Dropbox or Google Drive case study covering chunking, deduplication, a metadata database, block storage on S3, sync notifications, conflicts and versioning.35 min
    12. 12Design a Proximity ServiceA Yelp or nearby-places case study comparing geohash, quadtree and S2 for radius search, with read-heavy caching and a plan for updating business data.35 min
    13. 13Design a Ticket Booking SystemA case study of a BookMyShow or IRCTC style booking system, covering seat inventory, expiring holds, double-booking prevention and flash-sale traffic.40 min
    14. 14Design a Food Delivery SystemA case study of a Swiggy or Zomato style platform, covering restaurant search, cart and orders, the order state machine, partner assignment, live tracking and ETA.40 min
    15. 15Design a Real-Time LeaderboardA case study of a gaming leaderboard built on Redis sorted sets, covering rank queries, top-K reads, sharding past one node and daily or weekly boards.35 min
    16. 16Design an E-commerce CheckoutA case study of an Amazon style shop, covering catalog, cart, inventory reservation, the order and payment saga, idempotency and flash sales.40 min