Kubernetes questions, answered
Short answers to the questions we hear most often about Kubernetes and platform engineering.
Managed Kubernetes or run our own?
In public cloud, managed is almost always the right start. On-premises or with strict data rules, you run it yourself, ideally with a well-supported distribution and an upgrade routine.
What are the alternatives to Kubernetes?
Virtual machines with good automation (Ansible, Terraform), managed container services such as AWS ECS or Azure Container Apps, or a platform-as-a-service. For a handful of applications these are often simpler and cheaper.
Can databases run on Kubernetes?
Yes, with operators and good storage, and many teams do. For critical databases, many organisations still prefer managed databases or dedicated servers. Decide per database.
How many people does it take to run?
A small platform can be run by two or three people with good automation, especially on managed Kubernetes. Self-run, multi-cluster platforms need a dedicated team.
Does Kubernetes save money?
It can, by packing workloads more tightly and scaling down when idle. But it adds platform and people cost. The bigger gains are usually faster, safer delivery.
Where do we start?
With the fit check, then one lean cluster and two friendly teams.
More Kubernetes guides: Platform design · GitOps delivery · Security · Multi-cluster and on-prem · Kubernetes FAQ · Use case: festive-season scale
Planning a Kubernetes platform, or fixing one?
Tell us how many applications and teams you have, how you deploy today and where it should run. We will come back with an honest view: whether Kubernetes fits, what a lean platform looks like, and what it takes to run.