That 3AM Outage Could Have Been Avoided: An Introduction to Chaos Engineering

Most production outages are not surprises. They are failures we never thought to test for. Chaos engineering is the practice of finding those failures before they find you, and LitmusChaos is the open-source tool that makes it accessible to any team shipping software, not just the ones with a dedicated SRE team.

This talk covers the fundamentals of chaos engineering and LitmusChaos from the ground up. No prior experience needed.

What we will cover:

• Why production systems fail in ways CI/CD pipelines never catch

• What chaos engineering actually is and how it differs from load testing or stress testing

• How LitmusChaos works as a Kubernetes-native chaos orchestration platform

• A live walkthrough of an experiment: pod failure

• How to read experiment results and decide what to do next

LitmusChaos is a CNCF incubating project used in production by teams at Canonical, Red Hat, and Flipkart. The speaker is a maintainer and community manager for LitmusChaos, and will share real patterns seen across the adopter community, not just the happy path from the docs.

By the end of this talk, you will understand what chaos engineering is, why it matters for any team shipping software, and how to run your first experiment the same week.

• Chaos engineering is a proactive resilience practice, not a destructive one, and it is approachable for teams of any size

• LitmusChaos provides a Kubernetes-native way to define, run, and observe chaos experiments without putting production at risk

• You can run your first LitmusChaos experiment on a staging cluster within a day of watching this talk

• Reading chaos results is as important as running the experiments — the talk covers how to act on what you find