• Kubernetes is designed to make application highly available.
    • That is why there are multiple replications and essentially multiple applications running at the same time.
    • If one stops working, then at least two others are accepting traffic.
  • Pods are the Kubernetes resource that run our applications and microservices.
    • To make sure an applicaiton is highly available, is to organise your pods in a Kubernetes cluster.
  • Example deployment.yaml: 
    --- 
apiVersion: apps/v1
kind: Deployment
metadata:
name: pod-info-deployment
namespace: development
labels:
  app: pod-info
spec:
replicas: 1
selector:
  matchLabels:
    app: pod-info
template:
  metadata:
    labels:
      app: pod-info
  spec:
    containers:
    - name: pod-info-container
      image: kimschles/pod-info-app:latest
      ports:
      - containerPort: 3000
      env:
        - name: POD_NAME
          valueFrom:
            fieldRef:
              fieldPath: metadata.name
        - name: POD_NAMESPACE
          valueFrom:
            fieldRef:
              fieldPath: metadata.namespace
        - name: POD_IP
          valueFrom:
            fieldRef:
              fieldPath: status.podIP
  • apiVersion: apps/v1 –> This is the API group we are sending the request to.
  • kind: Deployment –> This is the kind of Kubernetes object we want to create.
  • The next section is metadata: 
    metadata:
name: pod-info-deployment
namespace: development
labels:
  app: pod-info
  • In the metadata section, the deployment name is pod-info-deployment . We are specifying that the pods should be in the development namespace. All pods in this group are labelled with the app name of pod-info.
  • To make the applications highly available.
    • Under spec, we can specify how many replicas of the particular container that we want to run.
  • At the following, we have: 
      spec:
    containers:
    - name: pod-info-container
      image: kimschles/pod-info-app:latest
      ports:
      - containerPort: 3000
  • The above are the specifics about the particular container that we want to run in the pod.
    • The name is pod-info-container
    • The image is from the authors Docker Hub Registry, which is kimschles/pod-info-app:latest.
    • The port for the traffic is going to be 3000, this is where the traffic is directed.
  • Then we have the environment variables of the container: 
    env:
        - name: POD_NAME
          valueFrom:
            fieldRef:
              fieldPath: metadata.name
        - name: POD_NAMESPACE
          valueFrom:
            fieldRef:
              fieldPath: metadata.namespace
        - name: POD_IP
          valueFrom:
            fieldRef:
              fieldPath: status.podIP
  • Get the list of namespaces with kubectl get ns
  • In this case, need to make sure the development namespace is applied first. This is again done with: 
    kubectl apply -f <yaml file>
  • Then we apply the deployment.yml file with: 
    kubectl apply -f deployment.yml
  • To list all of the deployments in the development namespace, we can do: 
    kubectl get deployments -n development
  • The -n is for namespace.
  • To check the pods that the delopment created, we can do: 
    kubectl get pods -n development
  • To delete a pod, we do: 
    kubectl delete pod pod-info-deployment-7587d5cc86-9cm4t -n development
  • The pod in the development namespace will now be deleted.
  • There will be a new pod created however, as the Kubernetes deployment will always create 3 pods.

Updated: