Exposing applications using services

This page shows how to create Kubernetes Services in a Google Kubernetes Engine (GKE) cluster. For an explanation of the Service concept and a discussion of the various types of Services, see Understand Kubernetes Services.

Introduction

The idea of a Service is to group a set of Pod endpoints into a single resource. You can configure various ways to access the grouping. By default, you get a stable cluster IP address that clients inside the cluster can use to contact Pods in the Service. A client sends a request to the stable IP address, and the request is routed to one of the Pods in the Service.

There are five types of Services:

  • ClusterIP (default)
  • NodePort
  • LoadBalancer
  • ExternalName
  • Headless

Autopilot clusters are public by default. If you opt for a private Autopilot cluster, you must configure Cloud NAT to make outbound internet connections, for example pulling images from DockerHub.

This topic has several exercises. In each exercise, you create a Deployment and expose its Pods by creating a Service. Then you send an HTTP request to the Service.

Before you begin

Before you start, make sure you have performed the following tasks:

  • Enable the Google Kubernetes Engine API.
    • Enable Google Kubernetes Engine API
  • If you want to use the Google Cloud CLI for this task, install and then initialize the gcloud CLI. If you previously installed the gcloud CLI, get the latest version by running gcloud components update.
* Create a GKE cluster.

Creating a Service of type ClusterIP

In this section, you create a Service of type ClusterIP.

kubectl apply

Here is a manifest for a Deployment:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: my-deployment
spec:
  selector:
    matchLabels:
      app: metrics
      department: sales
  replicas: 3
  template:
    metadata:
      labels:
        app: metrics
        department: sales
    spec:
      containers:
      - name: hello
        image: "us-docker.pkg.dev/google-samples/containers/gke/hello-app:2.0"

Copy the manifest to a file named my-deployment.yaml, and create the Deployment:

kubectl apply -f my-deployment.yaml

Verify that three Pods are running:

kubectl get pods

The output shows three running Pods:

NAME                            READY   STATUS    RESTARTS   AGE
my-deployment-dbd86c8c4-h5wsf   1/1     Running   0          7s
my-deployment-dbd86c8c4-qfw22   1/1     Running   0          7s
my-deployment-dbd86c8c4-wt4s6   1/1     Running   0          7s

Here is a manifest for a Service of type ClusterIP:

apiVersion: v1
kind: Service
metadata:
  name: my-cip-service
spec:
  type: ClusterIP
  # Uncomment the below line to create a Headless Service
  # clusterIP: None
  selector:
    app: metrics
    department: sales
  ports:
  - protocol: TCP
    port: 80
    targetPort: 8080

The Service has a selector that specifies two labels:

  • app: metrics
  • department: sales

Each Pod in the Deployment that you created previously has those two labels. So the Pods in the Deployment will become members of this Service.

Copy the manifest to a file named my-cip-service.yaml, and create the Service:

kubectl apply -f my-cip-service.yaml

Wait a moment for Kubernetes to assign a stable internal address to the Service, and then view the Service:

kubectl get service my-cip-service --output yaml

The output shows a value for clusterIP:

spec:
  clusterIP: 10.59.241.241

Make a note of your clusterIP value for later.

Console

Create a Deployment

  1. Go to the Workloads page in the Google Cloud console.

    Go to Workloads

  2. Click Deploy.

  3. Under Specify container, select Existing container image.

  4. For Image path, enter us-docker.pkg.dev/google-samples/containers/gke/hello-app:2.0

  5. Click Done, then click Continue.

  6. Under Configuration, for Application name, enter my-deployment.

  7. Under Labels, create the following labels:

    • Key: app and Value: metrics
    • Key: department and Value: sales

  8. Under Cluster, choose the cluster in which you want to create the Deployment.

  9. Click Deploy.

  10. When your Deployment is ready, the Deployment details page opens. Under Managed pods, you can see that your Deployment has one or more running Pods.

Create a Service to expose your Deployment

  1. On the Deployment details page, click Actions > Expose.

  2. In the Expose dialog, under Port mapping, set the following values:

    • Port: 80
    • Target port: 8080
    • Protocol: TCP

  3. From the Service type drop-down list, select Cluster IP.

  4. Click Expose.

  5. When your Service is ready, the Service details page opens, and you can see details about your Service. Under Cluster IP, make a note of the IP address that Kubernetes assigned to your Service. This is the IP address that internal clients can use to call the Service.

Accessing your Service

List your running Pods:

kubectl get pods

In the output, copy one of the Pod names that begins with my-deployment.

NAME                            READY   STATUS    RESTARTS   AGE
my-deployment-dbd86c8c4-h5wsf   1/1     Running   0          2m51s

Get a shell into one of your running containers:

kubectl exec -it POD_NAME -- sh

Replace POD_NAME with the name of one of the Pods in my-deployment.

In your shell, install curl:

apk add --no-cache curl

In the container, make a request to your Service by using your cluster IP address and port 80. Notice that 80 is the value of the port field of your Service. This is the port that you use as a client of the Service.

curl CLUSTER_IP:80

Replace CLUSTER_IP with the value of clusterIP in your Service.

Your request is forwarded to one of the member Pods on TCP port 8080, which is the value of the targetPort field. Note that each of the Service's member Pods must have a container listening on port 8080.

The response shows the output of hello-app:

Hello, world!
Version: 2.0.0
Hostname: my-deployment-dbd86c8c4-h5wsf

To exit the shell to your container, enter exit.

Creating a Service of type NodePort

In this section, you create a Service of type NodePort.

kubectl apply

Here is a manifest for a Deployment:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: my-deployment-50000
spec:
  selector:
    matchLabels:
      app: metrics
      department: engineering
  replicas: 3
  template:
    metadata:
      labels:
        app: metrics
        department: engineering
    spec:
      containers:
      - name: hello
        image: "us-docker.pkg.dev/google-samples/containers/gke/hello-app:2.0"
        env:
        - name: "PORT"
          value: "50000"

Notice the env object in the manifest. The env object specifies that the PORT environment variable for the running container will have a value of 50000. The hello-app application listens on the port specified by the PORT environment variable. So in this exercise, you are telling the container to listen on port 50000.

Copy the manifest to a file named my-deployment-50000.yaml, and create the Deployment:

kubectl apply -f my-deployment-50000.yaml

Verify that three Pods are running:

kubectl get pods

Here is a manifest for a Service of type NodePort:

apiVersion: v1
kind: Service
metadata:
  name: my-np-service
spec:
  type: NodePort
  selector:
    app: metrics
    department: engineering
  ports:
  - protocol: TCP
    port: 80
    targetPort: 50000

Copy the manifest to a file named my-np-service.yaml, and create the Service:

kubectl apply -f my-np-service.yaml

View the Service:

kubectl get service my-np-service --output yaml

The output shows a nodePort value:

...
  spec:
    ...
    ports:
    - nodePort: 30876
      port: 80
      protocol: TCP
      targetPort: 50000
    selector:
      app: metrics
      department: engineering
    sessionAffinity: None
    type: NodePort
...

Create a firewall rule to allow TCP traffic on your node port:

gcloud compute firewall-rules create test-node-port \
    --allow tcp:NODE_PORT

Replace NODE_PORT with the value of the nodePort field of your Service.

Console

Create a Deployment

  1. Go to the Workloads page in the Google Cloud console.

    Go to Workloads

  2. Click Deploy.

  3. Under Specify container, select Existing container image.

  4. For Image path, enter us-docker.pkg.dev/google-samples/containers/gke/hello-app:2.0.

  5. Click Add Environment Variable.

  6. For Key, enter PORT, and for Value, enter 50000.

  7. Click Done, then click Continue.

  8. Under Configuration, for Application name, enter my-deployment-50000.

  9. Under Labels, create the following labels:

    • Key: app and Value: metrics
    • Key: department and Value: engineering

  10. Under Cluster, choose the cluster in which you want to create the Deployment.

  11. Click Deploy.

  12. When your Deployment is ready, the Deployment details page opens. Under Managed pods, you can see that your Deployment has one or more running Pods.

Create a Service to expose your Deployment

  1. On the Deployment details page, click Actions > Expose.

  2. In the Expose dialog, under Port mapping, set the following values:

    • Port: 80
    • Target port: 50000
    • Protocol: TCP

  3. From the Service type drop-down list, select Node port.

  4. Click Expose.

  5. When your Service is ready, the Service details page opens, and you can see details about your Service. Under Ports, make a note of the Node Port that Kubernetes assigned to your Service.

Create a firewall rule for your node port

  1. Go to the Firewall policies page in the Google Cloud console.

    Go to Firewall policies

  2. Click Create firewall rule.

  3. For Name, enter test-node-port.

  4. From the Targets drop-down list, select All instances in the network.

  5. For Source IPv4 ranges, enter 0.0.0.0/0.

  6. Under Protocols and ports, select Specified protocols and ports.

  7. Select the tcp checkbox, and enter the node port value you noted.

  8. Click Create.

Get a node IP address

Find the external IP address of one of your nodes:

kubectl get nodes --output wide

The output is similar to the following:

NAME          STATUS    ROLES     AGE    VERSION        EXTERNAL-IP
gke-svc-...   Ready     none      1h     v1.9.7-gke.6   203.0.113.1

Not all clusters have external IP addresses for nodes. For example, if you have enabled private nodes, the nodes won't have external IP addresses.

Access your Service

In your browser's address bar, enter the following:

NODE_IP_ADDRESS:NODE_PORT

Replace the following:

  • NODE_IP_ADDRESS: the external IP address of one of your nodes, found when creating the service in the previous task.
  • NODE_PORT: your node port value.

The output is similar to the following:

Hello, world!
Version: 2.0.0
Hostname: my-deployment-50000-6fb75d85c9-g8c4f

Creating a Service of type LoadBalancer

In this section, you create a Service of type LoadBalancer.

kubectl apply

Here is a manifest for a Deployment:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: my-deployment-50001
spec:
  selector:
    matchLabels:
      app: products
      department: sales
  replicas: 3
  template:
    metadata:
      labels:
        app: products
        department: sales
    spec:
      containers:
      - name: hello
        image: "us-docker.pkg.dev/google-samples/containers/gke/hello-app:2.0"
        env:
        - name: "PORT"
          value: "50001"

Notice that the containers in this Deployment will listen on port 50001.

Copy the manifest to a file named my-deployment-50001.yaml, and create the Deployment:

kubectl apply -f my-deployment-50001.yaml

Verify that three Pods are running:

kubectl get pods

Here is a manifest for a Service of type LoadBalancer:

apiVersion: v1
kind: Service
metadata:
  name: my-lb-service
spec:
  type: LoadBalancer
  selector:
    app: products
    department: sales
  ports:
  - protocol: TCP
    port: 60000
    targetPort: 50001

Copy the manifest to a file named my-lb-service.yaml, and create the Service:

kubectl apply -f my-lb-service.yaml

When you create a Service of type LoadBalancer, a Google Cloud controller wakes up and configures an external passthrough Network Load Balancer. Wait a minute for the controller to configure the external passthrough Network Load Balancer and generate a stable IP address.

View the Service:

kubectl get service my-lb-service --output yaml

The output shows a stable external IP address under loadBalancer:ingress:

...
spec:
  ...
  ports:
  - ...
    port: 60000
    protocol: TCP
    targetPort: 50001
  selector:
    app: products
    department: sales
  sessionAffinity: None
  type: LoadBalancer
status:
  loadBalancer:
    ingress:
    - ip: 203.0.113.10

Console

Create a Deployment

  1. Go to the Workloads page in the Google Cloud console.

    Go to Workloads

  2. Click Deploy.

  3. Under Specify container, select Existing container image.

  4. For Image path, enter us-docker.pkg.dev/google-samples/containers/gke/hello-app:2.0.

  5. Click Add Environment Variable.

  6. For Key, enter PORT, and for Value, enter 50001.

  7. Click Done, then click Continue.

  8. Under Configuration, for Application name, enter my-deployment-50001.

  9. Under Labels, create the following labels:

    • Key: app and Value: products
    • Key: department and Value: sales

  10. Under Cluster, choose the cluster in which you want to create the Deployment.

  11. Click Deploy.

  12. When your Deployment is ready, the Deployment details page opens. Under Managed pods, you can see that your Deployment has one or more running Pods.

Create a Service to expose your Deployment

  1. On the Deployment details page, click Actions > Expose.

  2. In the Expose dialog, under Port mapping, set the following values:

    • Port: 60000
    • Target port: 50001
    • Protocol: TCP

  3. From the Service type drop-down list, select Load balancer.

  4. Click Expose.

  5. When your Service is ready, the Service details page opens, and you can see details about your Service. Under Load Balancer, make a note of the load balancer's external IP address.

Access your Service

Wait a few minutes for GKE to configure the load balancer.

In your browser's address bar, enter the following:

LOAD_BALANCER_ADDRESS:60000

Replace LOAD_BALANCER_ADDRESS with the external IP address of your load balancer.

The response shows the output of hello-app:

Hello, world!
Version: 2.0.0
Hostname: my-deployment-50001-68bb7dfb4b-prvct

Notice that the value of port in a Service is arbitrary. The preceding example demonstrates this by using a port value of 60000.

Creating a Service of type ExternalName

In this section, you create a Service of type ExternalName.

A Service of type ExternalName provides an internal alias for an external DNS name. Internal clients make requests using the internal DNS name, and the requests are redirected to the external name.

Here is a manifest for a Service of type ExternalName:

apiVersion: v1
kind: Service
metadata:
  name: my-xn-service
spec:
  type: ExternalName
  externalName: example.com

In the preceding example, the DNS name is my-xn-service.default.svc.cluster.local. When an internal client makes a request to my-xn-service.default.svc.cluster.local, the request gets redirected to example.com.

Using kubectl expose to create a Service

As an alternative to writing a Service manifest, you can create a Service by using kubectl expose to expose a Deployment.

To expose my-deployment, shown earlier in this topic, you could enter this command:

kubectl expose deployment my-deployment --name my-cip-service \
    --type ClusterIP --protocol TCP --port 80 --target-port 8080

To expose my-deployment-50000, show earlier in this topic, you could enter this command:

kubectl expose deployment my-deployment-50000 --name my-np-service \
    --type NodePort --protocol TCP --port 80 --target-port 50000

To expose my-deployment-50001, shown earlier in this topic, you could enter this command:

kubectl expose deployment my-deployment-50001 --name my-lb-service \
    --type LoadBalancer --port 60000 --target-port 50001

Cleaning up

After completing the exercises on this page, follow these steps to remove resources and prevent unwanted charges incurring on your account:

kubectl apply

Deleting your Services

kubectl delete services my-cip-service my-np-service my-lb-service

Deleting your Deployments

kubectl delete deployments my-deployment my-deployment-50000 my-deployment-50001

Deleting your firewall rule

gcloud compute firewall-rules delete test-node-port

Console

Deleting your Services

  1. Go to the Services page in the Google Cloud console.

    Go to Services

  2. Select the Services you created in this exercise, then click Delete.

  3. When prompted to confirm, click Delete.

Deleting your Deployments

  1. Go to the Workloads page in the Google Cloud console.

    Go to Workloads

  2. Select the Deployments you created in this exercise, then click Delete.

  3. When prompted to confirm, select the Delete Horizontal Pod Autoscalers associated with selected Deployments checkbox, then click Delete.

Deleting your firewall rule

  1. Go to the Firewall policies page in the Google Cloud console.

    Go to Firewall policies

  2. Select the test-node-port checkbox, then click Delete.

  3. When prompted to confirm, click Delete.

What's next