Overview:

• Services
• Microservices

Services

This is the part where you can access the service rather than pods. Pods can come and go. Example, you will create a deployment with 4 replicas. 1 pod was destroyed and new one was created so you need to access those pods dynamically.

Demo time!

So we will create a deployment first!

Running this command will:

• Create a deployment.
• A deployment that has 4 replicas/pods.
• Will run a container – a custom docker nginx container that just list the products (name and price) linuxacademycontent/store-products:1.0.0
• Each pods/replicas will run on port 80

cat << EOF | kubectl apply -f -
apiVersion: apps/v1
kind: Deployment
metadata:
  name: store-products
  labels:
    app: store-products
spec:
  replicas: 4
  selector:
    matchLabels:
      app: store-products
  template:
    metadata:
      labels:
        app: store-products
    spec:
      containers:
      - name: store-products
        image: linuxacademycontent/store-products:1.0.0
        ports:
        - containerPort: 80
EOF

Now it is time to create a Kubernetes service. The command below will:

• Create a Kubernetes service so that it allows you to dynamically access a group of replica pods
• That service store-product will run and listen on port 80
• It will target the pods that has the selector store-products (see screenshot above we got 4 pods with the pod name store-products)

cat << EOF | kubectl apply -f -
kind: Service
apiVersion: v1
metadata:
  name: store-products
spec:
  selector:
    app: store-products
  ports:
  - protocol: TCP
    port: 80
    targetPort: 80
EOF

Now you can access that specific web server IP from your master node machine to retrieve the products

curl 10.104.68.196

You can also access that web server service called store-products from a pod. Below is a demo where we run a command curl store-products inside the pod called busybox

kubectl exec busybox -- curl store-products

Microservices

Kubernetes Microservices is where small, independent services work together to form a whole application.

Example of small services:

• mysql – to store database like users,
• add to cart service (like a web code that process the add product to cart website)
• search service (like a web code that handles all the search queries)

Demo time!

We will use an existing and working YAML files to demo this Kubernetes microservices. This is from Linux Academy so we’ll download those YAML files.

The commands below will clone the robot shop store and make sure we are in our home directory.

cd ~/
git clone https://github.com/linuxacademy/robot-shop.git

This is what the robot shop directory looks like.

We’ll create a name space first so we can easily track/get info from this microservice.

kubectl create namespace robot-shop

Deploy the application objects (those YAML files from the screenshot above) to that name space robot-shop using the deployment descriptors from the git repo/robot-shop folder descriptor

Note:

I had to run these commands to tweak the robot-shop repo to fix the

# Replace apiVersion to apps/v1
sed -i 's/extensions\/v1beta1/apps\/v1/g' /root/robot-shop/K8s/descriptors/*.yaml   

# Add selector label
for FILE in /root/robot-shop/K8s/descriptors/*; do VAL=$(grep io.kompose.service $FILE | head -1 | awk '{print $2}');echo $VAL;sed -i "/replicas/i \  selector:\n\    matchLabels:\n\      io.kompose.service: $VAL" $FILE; done

Then run time to deploy the application objects to the namespace using the deployment descriptors from the Git repository

kubectl -n robot-shop create -f ~/robot-shop/K8s/descriptors/

Check the pods of this microservices deployment. You have to specify the name space robot-shop

kubectl get pods -n robot-shop -w