Why it matters
When a company runs dozens of containers across many servers, someone has to decide where each runs, restart them when they crash and add more when traffic spikes on salary day. Kubernetes does that. It appears in many Ghanaian DevOps adverts, mostly at mid and senior level, which is why this lesson is optional. Knowing the main ideas still helps you in interviews and in reading a team's setup on day one.
The core idea: desired state
You tell Kubernetes what you want ("three copies of the web app, always") and it works continuously to make reality match. If a container crashes or a server dies, Kubernetes notices the difference and starts a replacement. You describe; it maintains.
The main pieces
- Cluster: a group of machines run by Kubernetes.
- Control plane: the brain that stores desired state and makes decisions.
- Nodes: the worker machines that run your containers.
- Pod: the smallest unit, one or more containers that run together.
- Deployment: says which image to run and how many pods to keep.
- Service: a stable address that sends traffic to the right pods, even as pods come and go.