If you're using Kubernetes as containers orchestration, you get zero-downtime deployments of your workload for free. But is it exactly free? Can you be sure that you won't lose a single in-flight request while deploying? What about the changes in data and in the schema of your schema that you need to rollout? What if the deployment contains a bug that wasn't found in testing and is just about to hit all your users? What if your cluster nodes need to be maintained?
In this live-demo-driven session we'll have a look at all of these questions and what developers and operations have to take into account in order to deploy new versions of their software without disrupting a single user.
Комментарий программного комитета
Не всё то без downtime, что zero downtime. Достаем дьявола из деталей.