Putting Chaos Into Continuous Delivery to Increase Application Resiliency

Explore the integration of chaos engineering into continuous delivery pipelines to enhance application resiliency in this 27-minute conference talk from KubeCon + CloudNativeCon North America 2021. Learn how GitOps and chaos engineering are revolutionizing cloud-native continuous delivery practices. Discover techniques for implementing "chaos stages" in pre-production environments to assess SLO compliance. Gain insights from Juergen Etzlstorfer of Dynatrace and Karthik Satchitanand of Mayadata, maintainers of the CNCF sandbox projects Keptn and LitmusChaos, respectively. Understand how to construct pipelines that incorporate chaos experimentation, simulate real-world load, and implement quality gates to ensure resilient application deployment. Explore methods for integrating chaos tests into existing CD pipelines without rewriting them, all while adhering to GitOps principles. The presentation covers key topics such as the importance of chaos engineering, assumptions, churn, hypothesis formulation, and the use of LitmusChaos, concluding with a practical demonstration and essential takeaways.