Cianplan

Level 0: Orientation & prerequisites · L0.1 How the internet works & the design interview frame

Latency vs throughput

Distinguish latency and throughput and read p50/p99 at a basic level.

10 min · Free

What you take away

  • Latency is how long one request takes; throughput is how many requests finish per second. They are different dials.
  • More bandwidth or more servers raises throughput; neither shortens the distance a packet must travel.
  • Report percentiles, not averages: p50 is the typical user, p99 is the one who complains.
  • Concurrency = throughput × latency, which is how you turn a traffic estimate into a number of servers.

How the class is laid out

  1. The problem
  2. The simple version
  3. Why percentiles, and why the tail wins
  4. From traffic to servers
  5. Check your understanding