CSE Roadmap

Core tier · about 4 hours · 8 parts

Part 1 of 8 · 1 min read

Why this page comes sixth in Level 11

In this module you will

  • Scope a nearby-places service and estimate its search rate.
  • Explain why a plain latitude and longitude query is slow.
  • Describe the evenly divided grid, geohash and quadtree, and compare them.
  • Handle geohash boundary problems and searches that return too few results.
  • Scale the data, add caching, and answer the usual follow-ups.

Location search brings a new kind of data: two numbers that must be searched together. A normal index works in one dimension, so the design needs a way to turn a map into something an index can use. The same idea powers taxi dispatch, delivery and map apps. The book used is "System Design Interview, Volume 2" by Alex Xu, the proximity service chapter. Numbers marked "made up" only show the idea.