The design library
Building blocks, complete architectures, and variations worth comparing. Search by the problem you want to solve or a component you want to understand.
Classic designOfficial
Rank 200 million players in real time with sharded Redis sorted sets, count every game exactly once, and close a season fairly: one-hour boards for junior, senior and staff, with the theory behind them.
Classic designOfficial
Radius, k-nearest and map-box search over places that rarely move and drivers that ping every few seconds: R-trees, geohash, H3 and S2 cells, Redis GEO and OpenSearch. One-hour boards for junior, senior and staff, with the theory.
Classic designOfficial
An in-memory data store like Redis, from one event loop to Redis Cluster to a durable platform: one-hour boards for junior, senior and staff, with the theory behind them.
Classic designOfficial
Serve a product page a million times a second: a cheap query, replicas, Redis and the edge, with hot keys, stampedes and invalidation handled: one-hour boards for junior, senior and staff, with the theory behind them.
Classic designOfficial
Sell 10,000 units to a million people at noon with a lottery waiting room, a Redis gate before an Aurora ledger and holds that expire: one-hour boards for junior, senior and staff, with the theory behind them.
Classic designOfficial
Record runs and rides on the phone without signal, upload each once, show them to friends and rank athletes by week, month and year: one-hour boards for junior, senior and staff, with the theory behind them.
Classic designOfficial
Stacks of nearby people in under 300 ms, 2 billion swipes a day, and a match never missed even when both like at once: one-hour boards for junior, senior and staff, with the theory behind them.
Classic designOfficial
Search a billion posts a day by keyword, newest or most liked first, with an inverted index you build yourself: one-hour boards for junior, senior and staff, with the theory behind them.
Classic designOfficial
Photos and videos for 500 million daily users: presigned multipart uploads, a processing pipeline, CloudFront and a hybrid fan-out feed: one-hour boards for junior, senior and staff, with the theory behind them.
Classic designOfficial
A home feed for 2 billion users: fan-out on read, on write, and the hybrid for huge accounts, with precomputed feeds, cursors and hot posts: one-hour boards for junior, senior and staff, with the theory behind them.
Classic designOfficial
Price a trip, match a rider to a nearby driver who has ten seconds to accept, and never give one driver two rides: one-hour boards for junior, senior and staff, with the theory behind them.
Classic designOfficial
Run strangers' code safely in single-use microVMs, return verdicts within 5 seconds, and rank 100,000 contestants live: one-hour boards for junior, senior and staff, with the theory behind them.
Classic designOfficial
A terabyte of key-value pairs in memory across fifty nodes, read in a millisecond, with LRU, TTLs, replication, hot keys and stampedes handled: one-hour boards for junior, senior and staff, with the theory behind them.
Classic designOfficial
Allow or reject a million requests a second in under two milliseconds, across every gateway, with limits that hold and rules that change safely: one-hour boards for junior, senior and staff, with the theory behind them.
Classic designOfficial
The K most-viewed videos for the last hour, day, month and all time from 700,000 views a second, exactly and in milliseconds: one-hour boards for junior, senior and staff, with the theory behind them.
Building blockOfficial
Positions written to a cell-based index with a short expiry, so "who is near me" is a small read and anything that stops reporting disappears on its own.
Building blockOfficial
A bucket of tokens per key, refilled at a fixed rate, with refill-and-take done atomically in one place so every server agrees.
Building blockOfficial
Push new items into followers' timelines at write time, except for accounts with huge followings, whose posts are pulled and merged at read time.
Building blockOfficial
The service reads the cache first and fills it on a miss, so the cache holds only what is actually asked for and an outage degrades to slow rather than down.