We learned earlier that PODs without a Deployment object cannot be exposed as Services. In this section we will construct a YAML file that describes the Deployment and Service of the echoserver

K8 Objects via YAML file

Deployments

If you haven’t already created the k8_echoserver folder create a new folder called k8_echoserver. Navigate into the k8_echoserver folder.

Create a file called echoserver_deployment.yaml

Add the followings elements to the file

apiVersion: apps/v1
kind: Deployment
metadata:
  name: yml-echoserver-deployment
  namespace: default
spec:
  replicas: 1
  selector:
    matchLabels:
      service_id: echo-sever-from-yml
  template:
    metadata:
      labels:
        service_id: echo-sever-from-yml
    spec:
      containers:
      - name: yml-echo-server
        image: k8s.gcr.io/echoserver:1.4

Creating resources from the yaml file

Once you have created the file, from the shell where the file is located execute the command

kubectl apply -f echoserver_deployment.yaml

You should see a message that the Deployment has been created.

Visit the minikube dashboard, and look at the PODs, Deployments and Services, you will see the POD and the Deployment are there, but no service.

YAML Content Description

apiVersion - identifies the version of the schema the object should have

kind - a string that identifies the schema this object should have. Type kubectl api-resources to see the possible options.

metadata - Information that uniquely identifies the object

metadata.name - uniquely identifies this object within the current namespace

metadata.namespace - if not specified defaults to default

spec - What state do you desire for the object, in our example above we want it to have a

spec.replicas - target number of PODs
spec.selector.matchLabels - searches for a matching spec.template.metadata.labels value to determine which POD specification should be created

To remove the resources that have been created type

kubectl delete -f echoserver_deployment.yaml

Your Task

Use the steps from above and create a Deployment for the stream2stream/pingme image. It can be pulled from docker.io.