Ride Sharing HLD: Finding the Nearest Driver Without Asking All of Them
A ride sharing system design (Uber/Ola-style): the geospatial cell index that finds nearby drivers fast, the driver offer as a hold with a TTL, the double-match race, and surviving a hot region.
"Design Uber." You tap a pin on a map, and in the next few seconds the app says "finding your driver." Hold on those seconds — they're the whole interview. Somewhere out there are fifty thousand drivers scattered across the city, their dots crawling around the map as they drive. Your job, in a few milliseconds: find the handful closest to you, pick one, and make sure that the rider three blocks away who tapped at the same instant doesn't get offered the same car. The trap is that the obvious answer — "measure the distance from me to all fifty thousand drivers and take the smallest" — is a cata…
What’s inside
Read this one free
Sign in and your first premium article is on us — read Ride Sharing HLD: Finding the Nearest Driver Without Asking All of Them free.