Kubernetes

What it means

Kubernetes is an open-source system for automatically running, scaling and managing containerised applications across a group of servers called a cluster. It was originally developed by Google based on experience with its own infrastructure and is now maintained by the independent Cloud Native Computing Foundation.

Kubernetes builds on containerisation. While a container packages a single application, Kubernetes manages hundreds or thousands of containers: it decides which server they run on, restarts them on failure, moves them when a server goes down and adds instances when load rises.

The administrator does not describe individual steps but the desired state, for example three running instances of an application with a set amount of memory. The system continuously compares reality with that description and corrects it automatically. This delivers high availability without manual intervention.

How it works in practice

Kubernetes is used mainly for larger applications with a microservices architecture, where many parts run at once and load fluctuates. An online shop can automatically add capacity before Christmas and scale it down after the season, so it does not pay for servers it is not using.

Most companies do not run Kubernetes on their own servers but use a managed service from a major cloud provider. The provider looks after the control plane and updates, while the company handles only its applications. Deployment runs automatically through CI/CD within a DevOps culture.

Running on Kubernetes requires monitoring of performance, logs and costs. With poorly set limits, applications take more resources than they need and the monthly cloud bill grows. A regular review of consumption by application helps keep costs under control.

What to watch out for

Kubernetes is complex and requires specialist skills. For a small company with a simple website or internal system it is often overkill, bringing more work than benefit. Before adopting it, consider whether simpler hosting or a serverless service would be enough.

Cluster security depends on correctly configured access, networking and secrets. Mistakenly exposed interfaces are a common cause of data leaks. When running critical systems, cybersecurity requirements must also be considered, which for some companies follow from NIS2.