Deploy Multi-Tenant Applications using CAP Operator

CAP Operator provides a plugin to generate a Helm chart for your CAP application.

1. Add the CAP Operator plugin to your project by running the following command in your project root folder:

   bash

   Copy

   npm add @cap-js/cap-operator-plugin -D
   

2. Run the following command to generate a Helm chart for your CAP application:

   bash

   Copy

   npx cds add cap-operator --with-templates
   

   As a result, you see a newly created chart folder in your project. The chart folder holds the Helm configuration, including the values.yaml file where you add your container images.

3. Add your container image settings to your chart/values.yaml file:

   yaml

   Copy

   ...
   workloads:
       appRouter:
           ...
           deploymentDefinition:
               type: Router
               image: <your-container-registry>/incident-management-approuter:<image-version>
           ...
       server:
           ...
           deploymentDefinition:
               type: CAP
               image: <your-container-registry>/incident-management-srv:<image-version>
           ...
       contentDeploy:
           ...
           jobDefinition:
               type: Content
               image: <your-container-registry>/incident-management-html5-deployer:<image-version>
           ...
       tenantJob:
           ...
           jobDefinition:
               type: TenantOperation
               image: <your-container-registry>/incident-management-mtxs-sidecar:<image-version>
           ...
   ...
   

4. Add the EXIT_PROCESS_AFTER_UPLOAD environment variable to the content deploy job in your chart/values.yaml file to ensure that the HTML5 deployer exits after the upload is complete:

   yaml

   Copy

   ...
   contentDeploy:
       ...
       jobDefinition:
           type: Content
           image: <your-container-registry>/incident-management-html5-deployer:<image-version>
           env:
           - name: EXIT_PROCESS_AFTER_UPLOAD
             value: "true"
   ...
   

1. Run the following command to create a dedicated space for your application and enable the Istio service mesh to handle communication:

   bash

   Copy

   kubectl create namespace incident-management
   kubectl label namespace incident-management istio-injection=enabled
   

2. If you are using a private container registry, create a secret in the incident-management namespace so the cluster can pull your images. If your images are public, you can skip this step.

   bash

   Copy

   kubectl -n incident-management create secret generic regcred --from-file=.dockerconfigjson=$HOME/.docker/config.json --type=kubernetes.io/dockerconfigjson
   

3. Run the following command to get the cluster shoot domain:

   bash

   Copy

   kubectl get gateway -n kyma-system kyma-gateway -o jsonpath='{.spec.servers[0].hosts[0]}' | sed 's/^\*\.//'
   

   The result looks like this:

   bash

   Copy

   <xyz123>.kyma.ondemand.com
   

   <xyz123> is a placeholder for a string of characters that’s unique for your cluster.

4. Create a new file named trial-env.yaml in the project root folder with the following content and replace the placeholder values with your specific information:

   yaml

   Copy

   appName: <your-app-name>
   capOperatorSubdomain: cap-op
   clusterDomain: <cluster-shoot-domain> # Value obtained in the previous step
   globalAccountId: <your-global-account-id>
   providerSubaccountId: <your-provider-subaccount-id>
   providerSubdomain: <your-provider-subdomain>
   tenantId: <your-provider-tenant-id>
   imagePullSecret: regcred # Only include if you performed Step 2
   

   appName: Choose a name that is unique within your subaccount region. This prevents naming collisions with other deployments.

   capOperatorSubdomain: In Kyma clusters, CAP Operator subdomain default value is cap-op.

   clusterDomain: Use the domain string you retrieved in the previous step.

   globalAccountId: You can find this in the URL of your browser when you are viewing your subaccount in the SAP BTP Cockpit.

   

   providerSubaccountId, providerSubdomain and providerTenantId: In the SAP BTP cockpit, go to your subaccount Overview and check the General section. You can find all three values there.

   

   imagePullSecret: Only include this line if you are using a private registry. If you followed Step 2, set this to regcred.

5. To prepare your deployment, run the following command in your project root to generate the runtime-values.yaml file inside your chart folder:

   bash

   Copy

   npx cap-op-plugin generate-runtime-values --with-input-yaml trial-env.yaml
   

   This command maps your environment settings to the application’s configuration. It creates a runtime-values.yaml file that Helm uses during deployment to override the default settings in the values.yaml file with your specific cluster and account details.

6. Make sure that your SAP HANA Cloud instance is running. Free tier HANA instances are stopped overnight.

   Your SAP HANA Cloud service instance automatically stops overnight, according to the time zone of the region where the server is located. This means you need to restart your instance every day before you start working with it. You can restart your instance using the SAP BTP cockpit.

7. Deploy using the Helm command:

   bash

   Copy

   helm upgrade --install incident-management --namespace incident-management ./chart \
   --set-file serviceInstances.xsuaa.jsonParameters=xs-security.json -f ./chart/runtime-values.yaml
   

   This command installs the Helm chart from the chart folder with the release name incident-management in the incident-management namespace.

   With the helm upgrade –install command, you can install a new chart as well as upgrade an existing chart.

8. To check the status of your deployment:

   1. Open your Kyma dashboard.
   2. Choose Namespaces on the left and choose incident-management.
   3. Navigate to CAP Operator → CAP Application.
   4. You can see the status of your deployed application here. When the status is Consistent, your application is successfully deployed.
   

   In the example deployment, the unique appName is incident-tutorial.