Throughput vs. Latency: How To Debug A Latency Problem

Imagine you’re on call for the service you work on and you get paged in the middle of the night. Phone blaring, you stumble out of bed, sleepily turn on your computer, and look at your metrics. Looks like request latency went up a lot and users are getting errors. How do you figure out what’s going on? In this video, we’ll answer this question by looking at the relationship between throughput and latency, subtypes of latency like queuing latency, and techniques for debugging high latency. All you need is a little familiarity with software engineering. Let’s go!

00:00 Intro
00:28 Servers & Threads
00:53 What Is Latency?
01:24 What Is Throughput?
01:46 Ice Cream Analogy
02:19 Max Throughput
03:32 Queueing Latency
04:36 Server Simulation (don't miss this!)
07:17 Summary
07:58 How To Debug High Latency
08:33 Things Can Be Even More Complicated