Verify your deployment

To verify that your multicluster deployment works as intended, deploy a HelloWorld example service to your service mesh. Depending on your deployment model, you need to generate and send traffic within your mesh in a controlled way. This section includes instructions to verify that traffic reaches the clusters in your deployment following the configured behavior.

Before you begin

Ensure you have completed the following tasks:

• Verify you can access your clusters.

• Complete the initial configuration of your multicluster deployment.

Additionally, complete at least one of the following tasks once:

• Add a primary cluster
• Add a remote cluster

High-level flow

To verify your deployment, test the following scenarios as needed by your model:

• Services in different clusters respond to the appropriate requests.
• Services reach other services in the same cluster.
• Services can reach other services in a different cluster.

At a high-level, the verification flow consists of the following steps:

1. Deploy a service to both clusters.

   

2. Deploy different versions of the service to each cluster.

   

3. Deploy a load-generating service to both clusters.

   

4. Send traffic from the load-generating service in one cluster to the test service.

   

5. Send traffic from the load-generating service in the other cluster to the test service.

   

The following sections provide detailed procedures to complete each of the steps.

Deploy your test service

A client can call your HelloWorld endpoints if you configure DNS resolution for the helloworld hostname to an IP address. After you deploy the HelloWorld service to your mesh, Istio configures DNS for the helloworld hostname to resolve anywhere within the mesh. This guide uses the HelloWorld service as an example, but you can use your own service. To follow the best practice, create a namespace for your service before deploying it.

Using the previously set environment variables, deploy the HelloWorld service to Cluster_1 with the following steps:

1. Create the sample namespace in Cluster_1 with the following command:

   $ kubectl create --context=${CTX_1} namespace sample
   

2. Enable automatic sidecar injection for the sample namespace with the following command:

   $ kubectl label --context=${CTX_1} namespace sample \
       istio-injection=enabled
   

3. Deploy the HelloWorld service to Cluster_1 with the following command:

   $ kubectl create --context=${CTX_1} \
       -f ${ISTIO}/samples/helloworld/helloworld.yaml \
       -l app=helloworld -n sample
   

Using the previously set environment variables, deploy the HelloWorld service to Cluster_2 with the following steps:

1. Create the sample namespace in Cluster_2 with the following command:

   $ kubectl create --context=${CTX_2} namespace sample
   

2. Enable automatic sidecar injection for the sample namespace in Cluster_2 with the following command:

   $ kubectl label --context=${CTX_2} namespace sample \
       istio-injection=enabled
   

3. Deploy the HelloWorld service to Cluster_2 with the following command:

   $ kubectl create --context=${CTX_2} \
       -f ${ISTIO}/samples/helloworld/helloworld.yaml \
       -l app=helloworld -n sample
   

Deploy different service versions to different clusters

To verify cross-cluster load balancing, deploy v1 of the HelloWorld service to CLUSTER_1 and v2 to CLUSTER_2.

Using the previously set environment variables, deploy the different versions of the service to the clusters with the following steps:

1. Deploy v1 of the HelloWorld service to Cluster_1 with the following command:

   $ kubectl create --context=${CTX_1} \
       -f ${ISTIO}/samples/helloworld/helloworld.yaml \
       -l app=helloworld -l version=v1 -n sample
   

2. Deploy v2 of the HelloWorld service to Cluster_2 with the following command:

   $ kubectl create --context=${CTX_2} \
       -f ${ISTIO}/samples/helloworld/helloworld.yaml \
       -l app=helloworld -l version=v2 -n sample
   

3. Confirm the status v1 of the HelloWorld service with the following command:

   $ kubectl get pod --context=${CTX_1} -n sample
   NAME                            READY     STATUS    RESTARTS   AGE
   helloworld-v1-86f77cd7bd-cpxhv  2/2       Running   0          40s
   

4. Wait until the status of helloworld-v1 is Running.

5. Confirm the status v2 of the HelloWorld service with the following command:

   $ kubectl get pod --context=${CTX_2} -n sample
   NAME                            READY     STATUS    RESTARTS   AGE
   helloworld-v2-758dd55874-6x4t8  2/2       Running   0          40s
   

6. After the status of helloworld-v2 is Running, your cluster is ready to deploy a load-generating service.

Deploy a load-generating service

To generate the traffic needed to verify your deployment, deploy the sleep service to your clusters. We are using the sleep service as a simple load-generating service. If you are using your own service to verify your deployment, ensure that you have a traffic source creating a load for your services.

Using the previously set environment variables, deploy the load-generating service to the clusters with the following steps:

1. Deploy the sleep service to Cluster_1 with the following command:

   $ kubectl apply --context=${CTX_1} \
       -f ${ISTIO}/samples/sleep/sleep.yaml -n sample
   

2. Deploy the sleep service to Cluster_2 with the following command:

   $ kubectl apply --context=${CTX_2} \
       -f ${ISTIO}/samples/sleep/sleep.yaml -n sample
   

3. Confirm the status of the sleep service on Cluster_1 with the following command:

   $ kubectl get pod --context=${CTX_1} -n sample -l app=sleep
   sleep-754684654f-n6bzf           2/2     Running   0          5s
   

4. Wait until the status of the sleep service is Running.

5. Confirm the status of the sleep service on Cluster_2 with the following command:

   $ kubectl get pod --context=${CTX_2} -n sample -l app=sleep
   sleep-754684654f-dzl9j           2/2     Running   0          5s
   

6. After the status of the sleep service is Running, you can verify that cross-cluster load balancing is working as expected.

Verify cross-cluster load balancing

To verify that cross-cluster load balancing works as expected, call the HelloWorld service several times using the sleep service. To ensure load balancing is working properly, call the HelloWorld service from all clusters in your deployment.

Using the previously set environment variables, verify cross-cluster load balancing with following steps:

1. Send one request on port 5000 from the sleep service on Cluster_1 to the HelloWorld service with the following command:

   $ kubectl exec --context=${CTX_1} -it -n sample -c sleep \
       $(kubectl get pod --context=${CTX_1} -n sample -l \
           app=sleep -o jsonpath='{.items[0].metadata.name}') -- curl \
           helloworld.sample:5000/hello
   

2. To verify that the traffic reaches the endpoints of the HelloWorld service on your clusters, run the previous command multiple times.

3. If the endpoints on both clusters are responding, the output shows both v1 and v2 in the replies.

   Hello version: v2, instance: helloworld-v2-758dd55874-6x4t8
   Hello version: v1, instance: helloworld-v1-86f77cd7bd-cpxhv
   ...
   

4. Send a request on port 5000 from the sleep service on Cluster_2 to the HelloWorld service with the following command:

   $ kubectl exec --context=${CTX_2} -it -n sample -c sleep \
       $(kubectl get pod --context=${CTX_2} -n sample -l \
           app=sleep -o jsonpath='{.items[0].metadata.name}') -- curl \
           helloworld.sample:5000/hello
   

5. To verify that the traffic reaches the endpoints of the HelloWorld service on your clusters, run the previous command multiple times.

6. If the endpoints on both clusters are responding, the output shows both v1 and v2 in the replies.

   Hello version: v2, instance: helloworld-v2-758dd55874-6x4t8
   Hello version: v1, instance: helloworld-v1-86f77cd7bd-cpxhv
   ...
   

Congratulations!

You successfully verified the deployment of your multicluster mesh.

Clean up

After you no longer need your service mesh, you can remove Istio from your clusters.

Remove Istio with the following steps:

1. If you verified your deployment, delete the namespaces used in verification with the following command:

   $ kubectl delete namespace sample
   

2. Remove Istio from your clusters in the istio-system namespace with the following command:

   $ kubectl delete namespace istio-system
   

3. [Optional] Delete all files in your system’s working directory with the following command:

   $ rm -rf $WORK_DIR