Deploy and autoscale application

We are going to deploy the Hello world sample web application. This basic web application reads in an env variable TARGET and prints Hello ${TARGET}!. If TARGET is not specified, it will use World as the TARGET.

We will now deploy the application by specifying the image location and the TARGET env variable set to blue.

Knative defines a service.serving.knative.dev CRD to control the lifecycle of the application (not to be confused with kubernetes service). We will use the kn cli to create the Knative service: (This may take up to a minute)

kn service create demo --image gcr.io/knative-samples/helloworld-go --env TARGET=blue --autoscale-window 15s
```{{execute}}

We can now invoke the application using `curl`. We first need to figure out the IP address of minikube and the ingress port.

MINIKUBE_IP=$(minikube ip) INGRESS_PORT=$(kubectl get svc envoy –namespace contour-external –output ‘jsonpath={.spec.ports[?(@.port==80)].nodePort}')

Then invoke the application using curl:

curl http://$MINIKUBE_IP:$INGRESS_PORT/ -H ‘Host: demo.default.example.com’

### Scale down to zero
You can run `watch kubectl get pods`{{execute T2}} (may need two clicks) in a new Terminal tab to see a pod created to serve the requests. Knative will scale this pod down to zero if there are no incoming requests, we have configured this window to be 15 seconds above.

You can wait for the pods to scale down to zero and then issue the above `curl` again to see the pod spin up and serve the request.