Kubernetes: Up and Running, 3rd Edition

Chapter 21. Multicluster Application Deployments

Twenty chapters into this book, it should be clear that Kubernetes can be a complex topic, though of course we hope that if you have made it this far, it is less murky than it was. Given the complexities of building and running an application in a single Kubernetes cluster, why would you incur the added complexity of designing and deploying your application into multiple clusters?

The truth is that the demands of the real world mean that multicluster application deployment is a reality for most applications. There are many reasons for this, and it is likely that your application fits under at least one of these requirements.

The first requirement is one of redundancy and resiliency. Whether in the cloud or on-premise, a single datacenter is generally a single failure domain. Whether it is a hunter using a fiber-optic cable for target practice, a power outage from an ice storm, or simply a botched software rollout, any application deployed to a single location can fail completely and leave your users without recourse. In many cases, a single Kubernetes cluster is tied to a single location and thus is a single failure domain.

In some cases, especially in cloud environments, the Kubernetes cluster is designed to be regional. Regional clusters span across multiple independent zones and are thus resilient to the problems in the underlying infrastructure previously described. It would be tempting then to assume that such regional clusters ...