Scaling/rolling upgrades

For the adventurous you can edit deployments on the fly and modify the replicas field:

kubectl edit deployment veri-deployment

You can also edit the pod template and that will trigger the pods to be replaced, that might not happen in the nicest way though (unless you've defined a PodDisruptionBudget).

Instead of editing the manifests, Kubernetes ships with some built-in ways to scale and upgrade a Deployment, including things like HorizontalPodAutoscaler. Instead of using automation, we can scale the thing ourselves or do a rolling upgrade:

kubectl scale deployment/veri-deployment --replicas=3
kubectl rollout status deployment/veri-deployment
kubectl get pods

It might/might not be interesting to you that it's also possible to scale to 0, you can try that as well.

We can also change some of the fields such as env or image imperatively:

kubectl set env deployment/veri-deployment -e ECHO_TEXT="Some other text"
kubectl get deployment/veri-deployment -o yaml | grep -i "env\:" -A2

I think this latest change is actually causing lots of bugs, I'm sure, let's just undo the whole thing and move on:

kubectl rollout undo deployment/veri-deployment

In production clusters you would rarely use the kubectl set or kubectl rollout commands, but occasionally you might use if upgrading a cluster and bumping some of the image versions of kube-system DaemonSets for example. I'd recommend using GitOps controller such as ArgoCD to manage your application deployments.