Skip to content
All terms
Glossary

Kubernetes

A system that runs containers across multiple servers, monitors them, and restarts them on its own when they fail.

Kubernetes answers an operations question: how do you keep many containers running across several machines, distribute load, replace what has failed, and roll out new versions without interrupting the service? You describe the desired state and the system works continuously towards it.

That capability is impressive and still the wrong answer for many projects. Kubernetes brings complexity of its own - networking, storage, permissions, upgrades - and that wants running and understanding. A team operating one application on two servers solves the same problem with Docker Compose in a fraction of the time.

The turning point is where several teams need to ship independently, where load fluctuates heavily, or where resilience across multiple locations is required. Then the investment pays off, because the alternative would be many home-built individual solutions.

So we do not recommend for or against Kubernetes on principle but ask about the operating model: who will run this in two years, with what knowledge and what willingness? A platform nobody in the house commands is a risk, not a safeguard.

The service behind it

Security & Infrastructure