All problems

E-commerce Catalog

Browse and search a product catalog — and survive your own sale email.

medium~16 min Pro Simulator · +110 ptsTheory +45 pts
The build

Keep shoppers browsing through a seasonal sale while the search index slows and the catalog DB drops out.

Catalog browsing is read-heavy and wildly cacheable — the same bestseller pages over and over — while search queries need their own index off the browse path. Sized right, the CDN and product cache soak the sale surge and a DB outage never reaches a shopper; sized wrong, every page view is a DB read.

Components in play

  • Edge CDNServes product pages and images from the edge.
  • Catalog APIAssembles product pages; routes queries to the index.
  • Catalog dispatchSplits catalog traffic between the cache-backed browse path and the search index.
  • Product cacheHot products in memory so browsing skips the DB.
  • Catalog DBSource of truth for products, prices and stock.
  • Search indexFull-text queries — facets, typos, ranking — off the browse path.

Graded on these SLOs

  • ≥ 98.50%Browsing succeeds
  • ≤ 100 msPage p99
  • ≥ 9.0k rpsServe the sale
  • ≤ 85%Sale headroom
  • ≤ 8Lean footprint
Base 9,000 rps45s run 3 scripted failures

The brief

Design the product catalog for a storefront like Amazon. Shoppers browse product and category pages, and run full-text searches with facets and typo tolerance. Browsing is wildly read-heavy and cacheable — the same bestsellers over and over — while search is a different access pattern entirely and shouldn't compete with browse traffic on the same store. A seasonal sale can multiply browsing overnight, and the catalog DB will pick that moment to have a bad day.

Functional

  • Browse product and category pages
  • Full-text search with facets (price, brand, rating)
  • Update price/stock for a product

Non-functional

  • Read-heavy browsing (~50:1 over writes), low-latency page loads
  • Search must not degrade browse latency, or vice versa
  • High availability — checkout depends on the catalog staying up

Warm-up: design decisions

Optional theory to prime the calls a senior engineer would make before you build.

Pro

Solving is a Pro feature

Unlock the on-canvas AI design coach, per-run design reviews, the chaos simulator and all 12 problems for $15/month.

  • AI design coach on your canvas — pins failure points & over-engineering
  • Honest AI design review on every run
  • All 12 problems + unlimited chaos simulator
  • Points, levels & leaderboard
Get started