5 Ways to Improve Resilience in the Cloud

Your cloud system will fail. It's inevitable.

It even happens to the biggest tech companies.

In 2011, AWS suffered a major outage in one of its North Virginia availability zones, bringing down big names like Reddit and Quora. Amid the outage, one company managed to keep its services running: Netflix.

How did Netflix do it? They anticipated failure and built for it from the start. They had already tested their infrastructure’s resilience using a tool called Chaos Monkey, which randomly terminates instances in production to ensure the system can withstand instance failures without impacting customers.

This case study indicates resilience isn’t about luck—it’s engineered. And in a world increasingly dependent on the cloud, every engineer should know how to design for resilience.

Today, I'll cover:

• 5 proven techniques that drastically improve resiliency

• How to implement these strategies in major cloud providers: AWS, Azure, and GCP

• A 4-step framework to choose the right resiliency technique for your use case

Let’s get started.