This is the final blog in a series of posts that explore different options for running Java EE workloads on Azure. In this part, we will run the Java EE app on a Kubernetes cluster in Azure.

The previous parts covered how to deploy a Java EE application to an application server which is set up in a Virtual Machine on Microsoft Azureas well as a Docker container in Azure Container Instances.

The example used in the blog post is a simple three-tier application that uses Java EE 8 specifications such as JAX-RS, EJB, CDI, JPA, JSF, Bean Validation. We will use the Payara Server to deploy the application and use PostgreSQL as the relational database.

During the course of the tutorial, we will cover:

Postgres setup on Azure
Setup and configure Azure Kubernetes Service cluster and Azure Container Registry
Dockerize the Java EE app
Deploy the application to Kubernetes
Explore its functionality

Except for minor changes, the application used in this tutorial has been adapted from this project by Reza Rahman

Pre-requisites

You will need a Microsoft Azure account and the Azure CLI to work through the tutorial.

If you don’t have a Microsoft Azure account, go ahead and sign up for a free one!. The Azure CLI is a cross-platform command-line experience for managing Azure resources — please install it using these instructions.

First things first…

Set your Azure Subscription ID using the Azure CLI which will be used for this tutorial.

To set your Azure subscription ID

export AZURE_SUBSCRIPTION_ID=[to be filled]
az account set --subscription $AZURE_SUBSCRIPTION_ID

Create a resource group that will contain all the services (resources) which you will create as a part of this tutorial. A resource group is like a logical container that holds related resources for an Azure solution. The resource group includes those resources that you want to manage as a group.

To create a resource group

export AZURE_RESOURCE_GROUP_NAME=[to be filled]export AZURE_LOCATION=[to be filled]
az group create --name $AZURE_RESOURCE_GROUP_NAME --location $AZURE_LOCATION

Install Postgres on Azure

Azure Database for PostgreSQL is a relational database service based on the open-source Postgres database engine. It’s a fully managed database-as-a-service offering which is available in two deployment options, as a single server and as a Hyperscale (Citus) cluster

We will be using the single server option for the purposes of this tutorial

We will use the az postgres server createcommand to create a Postgres server instance on Azure. First, set up some of the server properties such as the name, admin user, etc.

export AZURE_POSTGRES_SERVER_NAME=[to be filled]
export AZURE_POSTGRES_ADMIN_USER=[to be filled]
export AZURE_POSTGRES_ADMIN_PASSWORD=[to be filled]
export SKU=B_Gen5_1
export STORAGE=5120

For storage and SKU options, please refer to the documentation

And, then invoke the command to initiate the database instance creation:

az postgres server create --resource-group $AZURE_RESOURCE_GROUP_NAME --name $AZURE_POSTGRES_SERVER_NAME  --location $AZURE_LOCATION --admin-user $AZURE_POSTGRES_ADMIN_USER --admin-password $AZURE_POSTGRES_ADMIN_PASSWORD --storage-size $STORAGE --sku-name $SKU

The provisioning process will take a few minutes.

To check the details of the Postgres database instance you just provisioned, invoke az postgres server show command

az postgres server show --resource-group $AZURE_RESOURCE_GROUP_NAME --name $AZURE_POSTGRES_SERVER_NAME

You should get a JSON response. Please note down the value for the fullyQualifiedDomainName attribute as you will be using this to connect to the Postgres instance later.

It should be of the format: *[AZURE_POSTGRES_DB_NAME].postgres.database.azure.com*

Azure Kubernetes Service (AKS) setup

You need the az aks create command to stand up a Kubernetes cluster on Azure

To keep things simple, the below command creates a single node cluster. Feel free to change the specification as per your requirements

export AKS_CLUSTER_NAME=[to be filled]az aks create --resource-group $AZURE_RESOURCE_GROUP --name $AKS_CLUSTER_NAME --node-count 1 --node-vm-size Standard_B2s --node-osdisk-size 30 --generate-ssh-keys

Get the AKS cluster credentials using az aks get-credentials - as a result, kubectl will now point to your new cluster. You can confirm the same

az aks get-credentials --resource-group $AZURE_RESOURCE_GROUP --name $AKS_CLUSTER_NAMEkubectl get nodes

If you are interested in learning Kubernetes and Containers using Azure, simply create a *free*account and get going! A good starting point is to use the quickstarts, tutorials and code samples in the documentation to familiarize yourself with the service. I also highly recommend checking out the 50 days Kubernetes Learning Path. Advanced users might want to refer to Kubernetes best practices or the watch some of the videos for demos, top features and technical sessions.

Allow AKS to access the Postgres database

The Postgres database is not accessible to external services by default. We can use the az postgres server firewall-rule create command to create a firewall rule to explicitly allow Azure services to access the Postgres instance. This will allow the JavaEE application deployed in AKS to communicate with Postgres.

export FIREWALL_RULE_NAME=AllowJavaEECafeAppOnAKSaz postgres server firewall-rule create --resource-group $AZURE_RESOURCE_GROUP_NAME --server-name $AZURE_POSTGRES_SERVER_NAME --start-ip-address=0.0.0.0 --end-ip-address=0.0.0.0 --name $FIREWALL_RULE_NAME

Note: This setting allows network connections from all IPs within the Azure network. For production use, try to configure the most restrictive firewall rules possible

Setup Azure Container Registry

Azure Container Registry is a managed, private Docker registry service to store and manage your private Docker container images (it based on the open-source Docker Registry 2.0). You can use Azure container registries with your existing container development and deployment pipelines, or use Azure Container Registry Tasks to build container images in Azure. You can either build on-demand, or fully automate builds with triggers such as source code commits and base image updates.

Let’s create a registry to store the Docker image for the JavaEE application. We will use the az acr createcommand

export ACR_NAME=[to-be-filled]az acr create --resource-group $AZURE_RESOURCE_GROUP_NAME --name $ACR_NAME --sku Basic --admin-enabled true

We are using the *Basic* SKU. Valid value are: *Basic*, *Classic*, *Premium*, *Standard*

You can log in to the registry once it’s created and check the login server

az acr login --name $ACR_NAME
az acr show --name $ACR_NAME --query loginServer --output table

You will use the ACR login server name soon. Its value follows the format: *[ACR_NAME].azurecr.io*

Configure Azure Container Registry to work with Azure Kubernetes Service

To access images stored in Azure Container Registry, you must grant the Azure Kubernetes Service service principal the correct rights to pull images from ACR.

Get the appId of the service principal which is associated with your AKS cluster

AKS_SERVICE_PRINCIPAL_APPID=$(az aks show --name $AKS_CLUSTER_NAME --resource-group $AZURE_RESOURCE_GROUP --query servicePrincipalProfile.clientId -o tsv)

Find the resource ID for Azure Container Registry

ACR_RESOURCE_ID=$(az acr show --resource-group $AZURE_RESOURCE_GROUP --name $ACR_NAME --query "id" --output tsv)

Grant acrpull permissions to AKS service principal

az role assignment create --assignee $AKS_SERVICE_PRINCIPAL_APPID --scope $ACR_RESOURCE_ID --role acrpull

Our AKS cluster along with ACR is ready to use!

Setup and prepare application image

Clone the git repository

git clone https://github.com/abhirockzz/javaee-on-azure-kubernetes
cd javaee-on-azure-kubernetes

You need to enter the Postgres connectivity information to the ``attribute of the section inweb.xml`.

You can find the *web.xml* file under *javaee-on-azure-iaas/src/main/webapp/WEB-INF*

The format is as follows:

jdbc:postgresql://[POSTGRES_FQDN]:5432/postgres?user=[AZURE_POSTGRES_ADMIN_USER]@[AZURE_POSTGRES_SERVER_NAME]&amp;password=[AZURE_POSTGRES_ADMIN_PASSWORD]&amp;sslmode=require

Here are the list placeholders which form a part of the JDBC URL:

POSTGRES_FQDN with value of fullyQualifiedDomainName
AZURE_POSTGRES_ADMIN_USER with admin user name used to provision PG
AZURE_POSTGRES_SERVER_NAME with server name used to provision PG
AZURE_POSTGRES_ADMIN_PASSWORD with admin password used to provision PG

Set the required values

export POSTGRES_FQDN=[to be filled]
export AZURE_POSTGRES_ADMIN_USER=[to be filled]
export AZURE_POSTGRES_SERVER_NAME=[to be filled]
export AZURE_POSTGRES_ADMIN_PASSWORD=[to be filled]

Simply use these commands to replace

export FILE_NAME=javaee-on-azure-iaas/src/main/webapp/WEB-INF/web.xmlsed -i 's/POSTGRES_FQDN/'"$POSTGRES_FQDN"'/g' $FILE_NAME
sed -i 's/AZURE_POSTGRES_SERVER_NAME/'"$AZURE_POSTGRES_SERVER_NAME"'/g' $FILE_NAME
sed -i 's/AZURE_POSTGRES_ADMIN_USER/'"$AZURE_POSTGRES_ADMIN_USER"'/g' $FILE_NAME
sed -i 's/AZURE_POSTGRES_ADMIN_PASSWORD/'"$AZURE_POSTGRES_ADMIN_PASSWORD"'/g' $FILE_NAME

Here is an e.g. of what the `` section will look like:

<data-source>
        <name>java:global/JavaEECafeDB</name>
        <class-name>org.postgresql.ds.PGPoolingDataSource</class-name>
        <url>jdbc:postgresql://foobar-pg.postgres.database.azure.com:5432/postgres?user=foobar@foobar-pg&amp;password=foobarbaz&amp;sslmode=require</url>
    </data-source>

The application is now configured. Let’s build it!

mvn clean install

You should have the WAR file available. To confirm

ls -lrt target | grep javaee-cafe.war

Build and push the image to Azure Container Registry

Our application artifact (WAR file) is ready. We can now build the Docker image and push it out to Azure Container Registry. Here is a quick look at the Dockerfile used for this

FROM payara/server-full
COPY target/javaee-cafe.war $DEPLOY_DIR
RUN wget https://jdbc.postgresql.org/download/postgresql-42.2.8.jar
RUN cp /opt/payara/postgresql-42.2.8.jar ${PAYARA_DIR}/glassfish/domains/${DOMAIN_NAME}/lib && rm /opt/payara/postgresql-42.2.8.jar
EXPOSE 8080

It builds on top of the base image for payara/server-full, copies the WARfile to a folder from where it can be automatically detected and deployed, downloads the Postgres JDBC driver and places it in the appropriate location for the Payara application server. That’s it!

export DOCKER_IMAGE=javaee-cafe
docker build -t $DOCKER_IMAGE .
docker tag $DOCKER_IMAGE $ACR_NAME.azurecr.io/$DOCKER_IMAGE

To push the image

docker push $ACR_NAME.azurecr.io/$DOCKER_IMAGE

For e.g., if the ACR_NAME (name of the Azure Container Registry) is javaeecafe-acr, the resulting Docker image will be javaeecafe-acr.azurecr.io/javaee-cafe

Use az acr repository list command to check the image.

az acr repository list --name $ACR_NAME --output table

Deploy the application to Azure Kubernetes Service

Before deploying the application, please update the Kubernetes manifest file javaee-cafe.yaml with the name of the Docker image. To be specific, update the spec.containers.image with the name of the Azure Container Registry which you specified above

It is assumed that the name of the Docker image is *javaee-azure*, if not please update that as well

spec:
  containers:
    - name: javaee-cafe
      image: <replace_me>.azurecr.io/javaee-azure

For e.g.

spec:
  containers:
    - name: javaee-cafe
      image: javaeecafe-acr.azurecr.io/javaee-azure

To deploy the application

kubectl apply -f javaee-cafe.yml

This should spin up a Pod. Wait for it to transition to Running state.

kubectl get pods -l=app=javaee-cafe -w

Once the Pod is Running, confirm that the application has been deployed successfully

kubectl logs -f <replace_with_pod_name>

The application deployment should be in progress and finally, you should see the logs similar to the one below (with the Successfully autodeployedmessage)

[AutoDeploy] Successfully autodeployed : /foo/bar/payara5/glassfish/domains/domain1/autodeploy/javaee-cafe.war.]]

Explore the application

We use a LoadBalancer Service type to ensure that our Java EE app is accessible outside of the cluster. The creation of a Kubernetes LoadBalancer``Service in Azure does exactly what it’s supposed to i.e. provision an Azure Load Balancer behind the scenes.

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

We can get the load balancer IP by using

kubectl get svc javaee-cafe

where *javaee-cafe* is the name of the *Service*

The value of the EXTERNAL-IP is the load balancer IP and will be used to access the application

Access the JSF front end

Use your browser to access http://[LOAD_BALANCER_IP]/javaee-cafe. You can use the UI to create, delete and see coffees.

Use the REST API

The application also exposes a REST API for creating, deleting and listing coffees.

export JAVAEE_AKS_REST=http://[LOAD_BALANCER_IP]/javaee-cafe/rest/coffees

e.g.

export JAVAEE_AKS_REST=http://23.101.24.139/javaee-cafe/rest/coffees

Create coffees

curl -X POST $JAVAEE_AKS_REST -d '{"name":"cappuccino","price":"10"}' -H "Content-Type: application/json"curl -X POST $JAVAEE_AKS_REST -d '{"name":"caffe-latte","price":"15"}' -H "Content-Type: application/json"

Get all coffees

curl -H "Accept: application/json" $JAVAEE_AKS_REST

You should see a JSON response listing both the coffee options you just added

Get a coffee by ID

curl -H "Accept: application/json" $JAVAEE_AKS_REST/1

Delete a coffee by ID

curl -X DELETE $JAVAEE_AKS_REST/1
curl -H "Accept: application/json" $JAVAEE_AKS_REST

Notice that cappuccino is now deleted

Scale

Right now, we have one instance of our application since we had set spec.replicas to 1 in the Kubernetes manifest. We can scale our application horizontally and Kubernetes will ensure that it spins up and maintains the required number of Pods. To add another instance

kubectl scale deployment javaee-cafe --replicas=2

To confirm that another Pod has been spun up:

kubectl get pods -l=app=javaee-cafe -w

You can continue accessing the application in the same manner and now the requests will be transparently load balanced amongst your app instances by the Load Balancer.

Clean up resources

Once you are done exploring the application, you can delete the resources. Since we used a resource group, it’s as easy as executing a single command.

Please be aware that this will delete all the resources in the group which includes the ones you created as part of the tutorial as well as any other service instances you might have if you used an already existing resource group

az group delete --name $AZURE_RESOURCE_GROUP_NAME

Summary

You learned how to leverage Docker containers to package your Java EE application and deploy it to a Kubernetes cluster in Azure along with a managed database offering for long term persistence.

That brings us to the end of this series exploring some of the common ways of deploying Java EE workloads to Azure. I hope you found it useful!

原文链接：https://medium.com/microsoftazure/deploying-java-ee-apps-to-azure-part-3-772e717bc4d1

创作场景

Deploying Java EE apps to Azure: Part 3