>_ IzzyLab

Topic breakdown

Replicas are identical pods of one Deployment. Several replicas share the load, and if one goes down, the rest keep working.

How many replicas to keep is recorded in the Deployment itself (the replicas field). You can change the number on the fly without recreating anything, with kubectl scale. For example, scale Deployment worker down to two pods:

kubectl scale deployment/worker --replicas=2
kubectl get deployment worker   # READY 2/2

Raise the number and Kubernetes creates the missing pods; lower it and it deletes the extra ones. kubectl rollout status deployment/<name> waits until the Deployment reaches the desired state.

What's checked

  1. Deployment catalog is set to 3 replicas
  2. All three replicas are ready

What it looks like