Multi-cluster Anthos Service Mesh using managed control plane

Diagram

Deploy Steps

1. Deploy the Terraform

    cd ~/src/anthos-examples/asm/infra
   

   update the values in var_inputs.auto.tfvars file

    terraform init && terraform apply
   

2. Enable multi-cluster mesh, from the cloud console, run:

    curl https://storage.googleapis.com/csm-artifacts/asm/asmcli_1.13 > asmcli
    chmod +x asmcli
    ./asmcli create-mesh <project_id> \
        <project_id>/us-west1/<project_id>-cluster1 \
        <project_id>/us-east1/<project_id>-cluster2
   

3. Get the cluster contexts

    export PROJECT_ID=<gcp_project_id>
    gcloud container clusters get-credentials ${PROJECT_ID}-cluster1 --region us-west1 --project ${PROJECT_ID}
    gcloud container clusters get-credentials ${PROJECT_ID}-cluster2 --region us-east1 --project ${PROJECT_ID}
   

4. On Cluster1, deploy ASM configuration and app manifests via kustomize

    cd anthos-examples/asm
    kubectl apply -k manifests/ --context=gke_${PROJECT_ID}_us-west1_${PROJECT_ID}-cluster1
   

5. On Cluster2, deploy only the app manifests via kustomize

    kubectl apply -k manifests/app/ --context=gke_${PROJECT_ID}_us-east1_${PROJECT_ID}-cluster2
   

Review your work

On Cluster1, you should have an ingress-gateway -> gateway -> virtual service -> service -> helloworld app v1 & v2

The ingress-gateway Service is created as a LoadBalancer, so there will be an external IP attached:

$> kubectl get svc -n asm-gateways --context=gke_${PROJECT_ID}_us-west1_${PROJECT_ID}-cluster1 
NAME                   TYPE           CLUSTER-IP     EXTERNAL-IP    PORT(S)                                      AGE
istio-ingressgateway   LoadBalancer   10.10.10.222   34.145.40.31   15021:30590/TCP,80:31782/TCP,443:31892/TCP   47h

Opening that IP in a browser (e.g. http://34.145.40.31/hello from the example output above) will show you the output of the helloworld app. Refreshing the page will give you a different pod id & version. With 4 pods running (2 per cluster, 2 per version) you will get a total of 4 different pod ids as you continue to refresh the page. This is because Istio is round-robin load-balancing the pods running behind the service.

On Cluster2, you should have gateway -> virtual service -> service -> helloworld app v1 & v2