1.7 KiB
| title | description |
|---|---|
| Minimal Viable Kubernetes | introducing minimal viable Kubernetes and its guiding principles |
Kubernetes is complicated but additional to this it is designed with cloud compute in mind.
Thus, self hosting options can be limiting.
Out of the box, Kubernetes 'expects' for there to be key ingredient of infrastructure to be present, for example, ingress, storage, load balancers. So if these fundamentals are not present, they need to be supplied in a form that Kubernetes can consume them.
Adding all these pre-requisites can be a challenge and even learning about Kubernetes, let alone setting up a development or 'lab' environment can quickly become messy. Knowing what to create in 'dev' as opposed to 'prod` may introduce unpredictable configuration, even technical debt.
Production environments can also become complex, difficult to manage, even over engineered which may add to technical debt.
Out of these scenarios, Minimal Viable Kubernetes (MVK) is a design pattern attempting to solve these issues.
MVK is there for a minimal setup to get a viable implementation of Kubernetes up and running initially for test and development but also for production use in an entirely self hosted environment, outside of any managed Kubernetes or cloud platform that is proprietary in its design and provisioning.
This increases portability, reduces the likelihood of 'vendor lock in'. It can reduce costs, provided that its users are or become more familiar with Kubernetes and its day to day operational and management requirements - spoiler alert, we think everyone should know more about this so as not to become totally reliant on others to do this for them - and above all, can increase control over our own digital sovereignty.