Enable Application Signals on a new Amazon EKS cluster with a sample app - Amazon CloudWatch
Step 1: Download the scriptsStep 2: Build and deploy the sample application
Services or capabilities described in Amazon Web Services documentation might vary by Region. To see the differences applicable to the China Regions, see Getting Started with Amazon Web Services in China (PDF).

Enable Application Signals on a new Amazon EKS cluster with a sample app

To try out CloudWatch Application Signals on a sample app before you instrument your own applications with it, follow the instructions in this section. These instructions use scripts to help you create an Amazon EKS cluster, install a sample application, and instrument the sample application to work with Application Signals.

The sample application is a Spring “Pet Clinic” application that is composed of four microservices. These services run on Amazon EKS on Amazon EC2 and leverage Application Signals enablement scripts to enable the cluster with the Java or Python auto-instrumentation agent.

Requirements

  • Currently, Application Signals monitors only Java and Python applications.

  • You must have the Amazon CLI installed on the instance. We recommend Amazon CLI version 2, but version 1 should also work. For more information about installing the Amazon CLI, see Install or update the latest version of the Amazon CLI.

  • The scripts in this section are intended to be run in Linux and macOS environments. For Windows instances, we recommend that you use an Amazon Cloud9 environment to run these scripts. For more information about Amazon Cloud9, see What is Amazon Cloud9?.

  • Install a supported version of kubectl. You must use a version of kubectl within one minor version difference of your Amazon EKS cluster control plane. For example, a 1.26 kubectl client works with Kubernetes 1.25, 1.26, and 1.27 clusters. If you already have an Amazon EKS cluster, you might need to configure Amazon credentials for kubectl. For more information, see Creating or updating a kubeconfig file for an Amazon EKS cluster.

  • Install eksctl. eksctl uses the Amazon CLI to interact with Amazon, which means it uses the same Amazon credentials as the Amazon CLI. For more information, see Installing or updating eksctl.

  • Install jq. jq is required to run the Application Signals enablement scripts. For more information, see Download jq.

Step 1: Download the scripts

To download the scripts to set up CloudWatch Application Signals with a sample app, you can download and uncompress the zipped GitHub project file to a local drive, or you can clone the GitHub project.

To clone the project, open a terminal window and enter the following Git command in a given working directory.

git clone https://github.com/aws-observability/application-signals-demo.git

Step 2: Build and deploy the sample application

To build and push the sample application images, follow these instructions.

Step 3: Deploy and enable Application Signals and the sample application

Be sure that you have completed the requirements listed in Enable Application Signals on a new Amazon EKS cluster with a sample app before you complete the following steps.

To deploy and enable Application Signals and the sample application

  1. Enter the following command. Replace new-cluster-name with the name that you want to use for the new cluster. Replace region-name with the name of the Amazon Region, such as us-west-1.

    This command sets up the sample app running in a new Amazon EKS cluster with Application Signals enabled. 

    # this script sets up a new cluster, enables Application Signals, and deploys the
# sample application
cd application-signals-demo/scripts/eks/appsignals/one-step && ./setup.sh new-cluster-name region-name

    The setup script takes about 30 minutes to run, and does the following:

    • Creates a new Amazon EKS cluster in the specified Region.

    • Creates the necessary IAM permissions for Application Signals (arn:aws:iam::aws:policy/AWSXrayWriteOnlyAccess and arn:aws:iam::aws:policy/CloudWatchAgentServerPolicy).

    • Enables Application Signals by installing the CloudWatch agent and Auto-instrumenting the sample application for CloudWatch metrics and X-Ray traces.

    • Deploys the PetClinic Spring sample application in the same Amazon EKS cluster.

    • Creates five CloudWatch Synthetics canaries, named pc-add-vist, pc-create-owners, pc-visit-pet, pc-visit-vet, pc-clinic-traffic. These canaries will run at a one-minute frequency to generate synthetic traffic for the sample app and demonstrate how Synthetics canaries appear in Application Signals.

    • Creates four service level objectives (SLOs) for the PetClinic application with the following names:

      • Availability for Searching an Owner

      • Latency for Searching an Owner

      • Availability for Registering an Owner

      • Latency for Registering an Owner

    • Creates the required IAM role with a custom trust policy granting Application Signals the following permissions:

      • cloudwatch:PutMetricData

      • cloudwatch:GetMetricData

      • xray:GetServiceGraph

      • logs:StartQuery

      • logs:GetQueryResults

  2. (Optional) If you want to review the source code for the PetClinic sample application, you can find them under the root folder.

    - application-signals-demo
  - spring-petclinic-admin-server
  - spring-petclinic-api-gateway
  - spring-petclinic-config-server
  - spring-petclinic-customers-service
  - spring-petclinic-discovery-server
  - spring-petclinic-vets-service
  - spring-petclinic-visits-service

  3. To view the deployed PetClinic sample application, run the following command to find the URL:

    kubectl get ingress

Step 4: Monitor the sample application

After completing the steps in the previous section to create the Amazon EKS cluster and deploy the sample application, you can use Application Signals to monitor the application.

Note

For the Application Signals console to start populating, some traffic must reach the sample application. Part of the previous steps created CloudWatch Synthetics canaries that generate traffic to the sample application.

Service health monitoring

After it is enabled, CloudWatch Application Signals automatically discovers and populates a list of services without requiring any additional setup.

To view the list of discovered services and monitor their health

  1. Open the CloudWatch console at https://console.amazonaws.cn/cloudwatch/.

  2. In the navigation pane, choose Application Signals, Services.

  3. To view a service, its operations, and its dependencies, choose the name of one of the services in the list.

    This unified, application-centric view helps provide a full perspective of how users are interacting with your service. This can help you triage issues if performance anomalies occur. For complete details about the Services view, see Monitor the operational health of your applications with Application Signals.

  4. Choose the Service Operations tab to see the standard application metrics for that service's operations. The operations are the API operations that the service calls, for example.

    Then, to view the graphs for a single operation of that service, choose that operation name.

  5. Choose the Dependencies tab to see the dependencies that your application has, along with the critical application metrics for each dependency. Dependencies include Amazon services and third-party services that your application calls.

  6. To view correlated traces from the service details page, choose a data point in one of the three graphs above the table. This populates a new pane with filtered traces from the time period. These traces are sorted and filtered based on the graph that you chose. For example, if you chose the Latency graph, the traces are sorted by service response time.

  7. In the CloudWatch console navigation pane, choose SLOs. You see the SLOs that the script created for the sample application. For more information about SLOs, see Service level objectives (SLOs).

(Optional) Step 5: Cleanup

When you're finished testing Application signals, you can use a script provided by Amazon to clean up and delete the artifacts created in your account for the sample application. To perform the cleanup, enter the following command. Replace new-cluster-name with the name of the cluster that you created for the sample app, and replace region-name with the name of the Amazon Region, such as us-west-1.

cd application-signals-demo/scripts/eks/appsignals/one-step && ./cleanup.sh new-cluster-name region-name