Premium Video

This video is available to premium subscribers only

Unlock Full Access

Every large system depends on components that can fail: servers crash, databases become unavailable, networks drop requests, and downstream services go down. A well-designed system expects these failures and continues working, or at least degrades gracefully, when one of them happens.

This lesson covers how to design systems that keep working when individual components fail, the main techniques used to contain those failures, and how to stop one broken service from taking down everything else.

It also covers how to approach failures in a system design interview, since for almost any design, you'll need to explain what happens when a part of it goes down.