Exposing the hello-minikube service fails to find a port

You get this error because you didn't set the container port from the command kubectl run hello-minikube --image=gcr.io/google_containers/echoserver:1.4 --port= as such the expose command doesn't know which container port to map to a node port and then error

You have to set exact container port as follow kubectl run hello-minikube --image=gcr.io/google_containers/echoserver:1.4 --port=80 assuming 80 is the port number and then run the expose again.

See bellow step by step of how I was able to replicate your error and then fix

C:\Users\innocent.anigbo\.minikube>kubectl run hello-kube --image=gcr.io/google_containers/echoserver:1.4 --port=deployment "hello-kube" createdC:\Users\innocent.anigbo\.minikube>kubectl get podsNAME                             READY     STATUS    RESTARTS   AGEhello-kube-1448409582-c9sm5      1/1       Running   0          1mhello-minikube-938614450-417hj   1/1       Running   1          8dhello-nginx-3322088713-c4rp4     1/1       Running   0          6mC:\Users\innocent.anigbo\.minikube>kubectl get deploymentNAME             DESIRED   CURRENT   UP-TO-DATE   AVAILABLE   AGEhello-kube       1         1         1            1           2mhello-minikube   1         1         1            1           8dhello-nginx      1         1         1            1           7mC:\Users\innocent.anigbo\.minikube>kubectl get serviceNAME          CLUSTER-IP   EXTERNAL-IP   PORT(S)        AGEhello-nginx   10.0.0.136   <nodes>       80:32155/TCP   4mkubernetes    10.0.0.1     <none>        443/TCP        20dC:\Users\innocent.anigbo\.minikube>kubectl expose deployment hello-kube --type=NodePorterror: couldn't find port via --port flag or introspectionSee 'kubectl expose -h' for help and examples.C:\Users\innocent.anigbo\.minikube>kubectl delete deployment hello-kubedeployment "hello-kube" deletedC:\Users\innocent.anigbo\.minikube>kubectl get podsNAME                             READY     STATUS    RESTARTS   AGEhello-minikube-938614450-417hj   1/1       Running   1          8dhello-nginx-3322088713-c4rp4     1/1       Running   0          11mC:\Users\innocent.anigbo\.minikube>kubectl run hello-kube --image=gcr.io/google_containers/echoserver:1.4 --port=80deployment "hello-kube" createdC:\Users\innocent.anigbo\.minikube>kubectl get podsNAME                             READY     STATUS    RESTARTS   AGEhello-kube-2715294448-0rxf2      1/1       Running   0          3shello-minikube-938614450-417hj   1/1       Running   1          8dhello-nginx-3322088713-c4rp4     1/1       Running   0          11mC:\Users\innocent.anigbo\.minikube>kubectl expose deployment hello-kube --type=NodePortservice "hello-kube" exposedC:\Users\innocent.anigbo\.minikube>kubectl get serviceNAME          CLUSTER-IP   EXTERNAL-IP   PORT(S)        AGEhello-kube    10.0.0.137   <nodes>       80:30004/TCP   3shello-nginx   10.0.0.136   <nodes>       80:32155/TCP   9mkubernetes    10.0.0.1     <none>        443/TCP        20d

Your pod is reappearing because deployment creates a replica set for the container which creates a new one if the current pod terminates.

Run this to check if there is a replica set deployed.

kubectl get rs

You should ideally delete the entire deployment

kubectl delete deployment <name>