cdCon+GitOpsCon 2023

How do you prevent a catastrophic meltdown? What happens to your customers' data when it happens? What is your blast radius? Mark's talk covers real world, advanced deployment techniques used to de-risk Red Hat's hosted OpenShift product. A shared control plane might make things faster and less expensive for customers, but it also concentrates risk to one area. If this blows up, a lot of unhappy customers would lose their data. Instead of trying to prevent meltdowns, this talk covers advanced deployment techniques that allow you to control the explosion, mitigate risk, and handle failure gracefully. Advanced techniques include canary instances, deployment rings, progressive rollout through sectors, feature flags, and automated ci/cd testing and promotions.