How to Integrate Microsoft Azure Cognitive Services with Kubernetes

Set Up Your Azure Cognitive Services

• Login to the [Azure Portal](https://portal.azure.com/).

 

• Navigate to the Azure Cognitive Services and create a new service instance for the specific service you are interested in, such as Computer Vision, Text Analytics, etc.

 

• Once created, note down the keys and the endpoint URL provided in the 'Keys and Endpoint' section.

Install and Configure the Azure CLI

• Ensure you have the Azure CLI installed on your local machine. If not, [download and install it](https://docs.microsoft.com/en-us/cli/azure/install-azure-cli).

 

• Log in to Azure using the CLI with the following command:

az login

• Configure your default subscription by running:

az account set --subscription "your-subscription-name"

Set Up a Kubernetes Cluster

• You can create a Kubernetes cluster using Azure Kubernetes Service (AKS). First, install the kubectl CLI tool if it's not already installed using:

az aks install-cli

• Now, create a resource group:

az group create --name myResourceGroup --location eastus

• Create an AKS cluster:

az aks create --resource-group myResourceGroup --name myAKSCluster --node-count 1 --enable-addons monitoring --generate-ssh-keys

• Get the kubectl context for the created cluster:

az aks get-credentials --resource-group myResourceGroup --name myAKSCluster

Deploy Your Application to Kubernetes

• Create a Kubernetes secret to store your Azure Cognitive Services keys for secure access:

kubectl create secret generic azure-cognitive-secret --from-literal=apiKey=YOUR_API_KEY --from-literal=endpoint=YOUR_ENDPOINT_URL

• Write a Kubernetes deployment YAML manifest file to deploy your application. Reference the secret in your application as environment variables:

apiVersion: apps/v1  
kind: Deployment  
metadata:  
  name: cognitive-service-app  
spec:  
  replicas: 1  
  selector:  
    matchLabels:  
      app: cognitive-service-app  
  template:  
    metadata:  
      labels:  
        app: cognitive-service-app  
    spec:  
      containers:  
      - name: cognitive-service-container  
        image: your-docker-image  
        env:  
        - name: API_KEY  
          valueFrom:  
            secretKeyRef:  
              name: azure-cognitive-secret  
              key: apiKey  
        - name: ENDPOINT  
          valueFrom:  
            secretKeyRef:  
              name: azure-cognitive-secret  
              key: endpoint

• Apply the deployment configuration:

kubectl apply -f deployment.yaml

• Verify that the pod is running with:

kubectl get pods

Expose Your Service

• Create a service to expose your deployment. You can choose to expose it as a LoadBalancer, NodePort, or ClusterIP. Here’s an example of a LoadBalancer service manifest:

apiVersion: v1  
kind: Service  
metadata:  
  name: cognitive-service  
spec:  
  type: LoadBalancer  
  ports:  
    - port: 80  
      targetPort: 8080  
  selector:  
    app: cognitive-service-app

• Apply the service configuration:

kubectl apply -f service.yaml

• Check the external IP of the service using:

kubectl get svc cognitive-service

Test Your Application

• Access the service using the external IP obtained in the previous step and verify the integration of Azure Cognitive Services with your application running on Kubernetes.

 

• Ensure your application is able to interact with the Azure Cognitive Services endpoints securely using the stored API keys.