Set up Google Kubernetes Engine Pods using manual Envoy injection

This guide shows you how to configure Google Kubernetes Engine or Kubernetes Pod hosts and the load balancing components that Cloud Service Mesh requires.

Before you follow the instructions in this guide, review Preparing for Cloud Service Mesh setup and make sure that you have completed the prerequisites.

You can configure Cloud Service Mesh using the Compute Engine load balancing SDK or REST APIs. See the load balancing API and gcloud references.

Configuring GKE/Kubernetes clusters for Cloud Service Mesh

This section describes the required steps to enable GKE/Kubernetes clusters to work with Cloud Service Mesh.

Creating the GKE cluster

GKE clusters must meet the following requirements:

The following example shows how to create a GKE cluster called traffic-director-cluster in the us-central1-a zone.


To create a cluster using Google Cloud console, perform the following steps:

  1. Go to the Kubernetes Engine menu in Google Cloud console.

    Go to the Google Kubernetes Engine menu

  2. Click Create cluster.

  3. Complete the following fields:

    • Name: Enter traffic-director-cluster.
    • Location type: Zonal.
    • Zone: us-central1-a.
  4. From the navigation pane, under Node Pools, click default-pool.

  5. The Size field indicates the number of nodes to create in the cluster. You must have available resource quota for the nodes and their resources (such as firewall routes).

  6. From the navigation pane, under default-pool, click Nodes.

  7. The Machine type field indicates the Compute Engine machine type to use for the instances. Each machine type is billed differently. For machine type pricing information, refer to the Compute Engine pricing page.

  8. From the navigation pane, under default-pool, click Security.

  9. Under Access scopes click Allow full access to all Cloud APIs.

  10. Customize your cluster as necessary.

  11. Click Create.

After you create a cluster in Google Cloud console, you need to configure kubectl to interact with the cluster. To learn more, refer to Generating a kubeconfig entry.


gcloud container clusters create traffic-director-cluster \
  --zone us-central1-a \
  --scopes= \

Obtaining the required GKE cluster privileges

For GKE, switch to the cluster(2) you just created by issuing the following command. This points kubectl to the correct cluster.

gcloud container clusters get-credentials traffic-director-cluster \
    --zone us-central1-a

Configuring GKE/Kubernetes services

This section shows how to prepare Kubernetes deployment specifications to work with Cloud Service Mesh. This consists of configuring services with NEGs as well as injecting sidecar proxies into Pod s that require access to the services managed by Cloud Service Mesh.

Configure the firewall rule

To verify that the backend Pods are running, you must configure a firewall rule allowing the health checker IP address ranges.


  1. Go to the Firewall policies page in the Google Cloud console.
    Go to the Firewall policies page
  2. Click Create firewall rules.
  3. On the Create a firewall rule page, supply the following information:
    • Name: Provide a name for the rule. For this example, use fw-allow-health-checks.
    • Network: Choose a VPC network.
    • Priority: Enter a number for the priority. Lower numbers have higher priorities. Be sure that the firewall rule has a higher priority than other rules that might deny ingress traffic.
    • Direction of traffic: Choose ingress.
    • Action on match: Choose allow.
    • Targets: Choose All instances in the network.
    • Source filter: Choose the correct IP range type.
    • Source IP ranges:,
    • Destination filter: Select the IP type.
    • Protocols and ports: Click Specified ports and protocols, then check tcp. TCP is the underlying protocol for all health check protocols.
    • Click Create.


  1. Use the following gcloud command to create a firewall rule named fw-allow-health-checks that allows incoming connections to instances in your network with the allow-health-checks tag. Replace NETWORK_NAME with the name of your network.

    gcloud compute firewall-rules create fw-allow-health-checks \
        --network NETWORK_NAME \
        --action ALLOW \
        --direction INGRESS \
        --source-ranges, \
        --rules tcp

For more information, see configure firewal rule for health checks.

Configuring GKE / Kubernetes services with NEGs

GKE services must be exposed through network endpoint groups (NEGs) so that you can configure them as backends of a Cloud Service Mesh backend service. Add the NEG annotation to your Kubernetes service specification and choose a name (by replacing NEG-NAME in the sample below) so that you can find it easily later. You need the name when you attach the NEG to your Cloud Service Mesh backend service. For more information on annotating NEGs, see Naming NEGs.

  annotations: '{"exposed_ports": {"80":{"name": "NEG-NAME"}}}'
  - port: 80
    name: service-test
    protocol: TCP
    targetPort: 8000

For each service, a standalone NEG is created, containing endpoints that are the pod's IP addresses and ports. For more information and examples, refer to Standalone network endpoint groups.

For demonstration purposes, you can deploy a sample service that serves its hostname over HTTP on port 80:

wget -q -O - \ \
| kubectl apply -f -

Verify that the new service hostname is created and the application Pod is running:

kubectl get svc

This returns:

service-test     ClusterIP   none          80/TCP    41m

kubectl get pods

This returns:

NAME                        READY     STATUS    RESTARTS   AGE
app1-6db459dcb9-zvfg2       1/1       Running   0          6m

Saving the NEG's name

Find the NEG created from the example above and record the NEG's name.


To view a list of network endpoint groups, go to the Network Endpoint Groups page in the Google Cloud console.
Go to the Network Endpoint Groups page


gcloud compute network-endpoint-groups list

This returns the following:

NAME                 LOCATION          ENDPOINT_TYPE      SIZE
NEG-NAME           us-central1-a     GCE_VM_IP_PORT      1

Save the NEG name in the NEG_NAME variable, for example:

NEG_NAME=$(gcloud compute network-endpoint-groups list \
| grep service-test | awk '{print $1}')

Configuring Google Cloud load balancing components for Cloud Service Mesh

The instructions in this section ensure that GKE services are accessible on the service VIP load balanced by Cloud Service Mesh, using a load balancing configuration similar to other Google Cloud Load Balancing products.

You must configure the following components:

The Cloud Service Mesh configuration example that follows makes these assumptions:

  1. The NEGs and all other resources are created in default network, with auto mode, in the zone us-central1-a.
  2. The NEG name for the cluster is stored in the ${NEG_NAME} variable.

Creating the health check

Create the health check.


  1. Go to the Health checks page in the Google Cloud console.
    Go to the Health checks page
  2. Click Create Health Check.
  3. For the name, enter td-gke-health-check.
  4. For the protocol, select HTTP.
  5. Click Create.


gcloud compute health-checks create http td-gke-health-check \

Creating the backend service

Create a global backend service with a load balancing scheme of INTERNAL_SELF_MANAGED. In the Google Cloud console, the load balancing scheme is set implicitly. Add the health check to the backend service.


  1. Go to the Cloud Service Mesh page in the Google Cloud console.

    Go to the Cloud Service Mesh page

  2. On the Services tab, click Create Service.

  3. Click Continue.

  4. For the service name, enter td-gke-service.

  5. Under Backend type, select Network endpoint groups.

  6. Select the network endpoint group you created.

  7. Set the Maximum RPS to 5.

  8. Click Done.

  9. Under Health check, select td-gke-health-check, which is the health check you created.

  10. Click Continue.


  1. Create the backend service and associate the health check with the backend service.

    gcloud compute backend-services create td-gke-service \
     --global \
     --health-checks td-gke-health-check \
     --load-balancing-scheme INTERNAL_SELF_MANAGED
  2. Add the backend NEGs to the backend service.

    gcloud compute backend-services add-backend td-gke-service \
     --global \
     --network-endpoint-group ${NEG_NAME} \
     --network-endpoint-group-zone us-central1-a \
     --balancing-mode RATE \
     --max-rate-per-endpoint 5

Creating the routing rule map

Use these instructions to create the route rule, forwarding rule, and internal IP address for your Cloud Service Mesh configuration.

Traffic sent to the internal IP address is intercepted by the Envoy proxy and sent to the appropriate service according to the host and path rules.

The forwarding rule is created as a global forwarding rule with the load-balancing-scheme set to INTERNAL_SELF_MANAGED.

You can set the address of your forwarding rule to If you do, traffic is routed based on the HTTP hostname and path information configured in the URL map, regardless of the actual IP address that the hostname resolves to. In this case, the URLs (hostname plus URL path) of your services, as configured in the host rules, must be unique within your service mesh configuration. That is, you cannot have two different services, with different set of backends, that both use the same hostname and path combination.

Alternatively, you can enable routing based on the actual destination VIP of the service. If you configure the VIP of your service as an address parameter of the forwarding rule, only requests destined to this IP address are routed based on the HTTP parameters specified in the URL map.


In the console, the target proxy is combined with the forwarding rule. When you create the forwarding rule, Google Cloud automatically creates a target HTTP proxy and attaches it to the URL map.

The route rule consist of the forwarding rule and the host and path rules (also known as the URL map).

  1. Go to the Cloud Service Mesh page in the Google Cloud console.

    Go to the Cloud Service Mesh page

  2. Click Routing rule maps

  3. Click Create Routing Rule.

  4. Enter td-gke-url-map as the Name of the URL map.

  5. Click Add forwarding rule.

  6. For the forwarding rule name, enter td-gke-forwarding-rule.

  7. Select your network.

  8. Select your Internal IP.

  9. Click Save.

  10. Optionally, add custom host and path rules or leave the path rules as the defaults.

  11. Set the host to service-test.

  12. Click Save.


  1. Create a URL map that uses the backend service.

    gcloud compute url-maps create td-gke-url-map \
       --default-service td-gke-service
  2. Create a URL map path matcher and a host rule to route traffic for your service based on hostname and a path. This example uses service-test as the service name and a default path matcher that matches all path requests for this host (/*). service-test is also the configured name of the Kubernetes service that is used in the sample configuration above.

    gcloud compute url-maps add-path-matcher td-gke-url-map \
       --default-service td-gke-service \
       --path-matcher-name td-gke-path-matcher
    gcloud compute url-maps add-host-rule td-gke-url-map \
       --hosts service-test \
       --path-matcher-name td-gke-path-matcher
  3. Create the target HTTP proxy.

    gcloud compute target-http-proxies create td-gke-proxy \
       --url-map td-gke-url-map
  4. Create the forwarding rule.

    gcloud compute forwarding-rules create td-gke-forwarding-rule \
      --global \
      --load-balancing-scheme=INTERNAL_SELF_MANAGED \
      --address= \
      --target-http-proxy=td-gke-proxy \
      --ports 80 --network default

At this point, Cloud Service Mesh is configured to load balance traffic for the services specified in the URL map across backends in the network endpoint group.

Depending on how your microservices are distributed on your network, you might need to add more forwarding rules or more host and path rules to the URL map.

Verifying the configuration by deploying a sample client for tests

This section shows how to reach Cloud Service Mesh backends from a client application.

To demonstrate functionality, you can deploy a sample Pod running Busybox. The pod has access to service-test, which was created in the previous section and receives traffic that is load balanced by Cloud Service Mesh.

Injecting a sidecar proxy into GKE / Kubernetes Pods

To access a service managed by Cloud Service Mesh, a Pod must have an xDS API-compatible sidecar proxy installed.

In this example, you deploy a Busybox client with an Istio-proxy sidecar and init containers added to the deployment using the reference specification.

If you are using the older APIs, replace the PROJECT_NUMBER and NETWORK_NAME variables with your project number and network name:

wget -q -O -
sed -i "s/PROJECT_NUMBER/PROJECT_NUMBER/g" trafficdirector_client_sample_xdsv3.yaml
sed -i "s/NETWORK_NAME/NETWORK_NAME/g" trafficdirector_client_sample_xdsv3.yaml
kubectl apply -f trafficdirector_client_sample_xdsv3.yaml

If you are using the new service routing APIs, which are currently in preview, replace the PROJECT_NUMBER and MESH_NAME variables with the project number and Mesh name:

wget -q -O -
sed -i "s/PROJECT_NUMBER/PROJECT_NUMBER/g" trafficdirector_client_new_api_sample_xdsv3.yaml
sed -i "s/MESH_NAME/MESH_NAME/g" trafficdirector_client_new_api_sample_xdsv3.yaml
kubectl apply -f trafficdirector_client_new_api_sample_xdsv3.yaml

The Busybox Pod has two containers running. The first container is the client based on the Busybox image and the second container is the Envoy proxy injected as a sidecar. You can get more information about the Pod by running the following command:

kubectl describe pods -l run=client

Reaching the backend service

Once configured, applications on Pods that have a sidecar proxy injected can access services managed by Cloud Service Mesh services. To verify the configuration, you can access a shell on one of the containers.

If you used the demo configuration provided in this guide, you can execute the following verification command to make sure that the hostname of the serving Pod is returned.

# Get name of the Pod  with busybox.
BUSYBOX_POD=$(kubectl get po -l run=client -o=jsonpath='{.items[0]}')

# Command to execute that tests connectivity to the service service-test.
TEST_CMD="wget -q -O - service-test; echo"

# Execute the test command on the Pod .
kubectl exec -it $BUSYBOX_POD -c busybox -- /bin/sh -c "$TEST_CMD"

Understanding traffic interception by the sidecar proxy

Note that, in this example, when the Busybox client makes requests to the backend service, each request is proxied by the sidecar proxy.

This demonstration application uses the Envoy proxy. Because of that, the client sees 'server: envoy' in the header of server responses.

To confirm this, use the following commands:

# Get the name of the Pod  with Busybox.
BUSYBOX_POD=$(kubectl get po -l run=client -o=jsonpath='{.items[0]}')

# Command to send a request to service-test and output server response headers.
TEST_CMD="wget -S --spider service-test; echo"

# Execute the test command on the Pod .
kubectl exec -it $BUSYBOX_POD -c busybox -- /bin/sh -c "$TEST_CMD"

In this example, you created a forwarding rule using the VIP address This means that Cloud Service Mesh forwards requests to the backend based on the Host header only. In this case, the destination IP address can be any address as long as the request host header matches the host defined in the URL map service-test.

To confirm that, run the following test commands:

# Get name of the Pod  with Busybox.
BUSYBOX_POD=$(kubectl get po -l run=client -o=jsonpath='{.items[0]}')

# Command to send a request to service-test setting the Host header and using a random IP address.
TEST_CMD="wget -q --header 'Host: service-test' -O -; echo"

# Execute the test command on the Pod .
kubectl exec -it $BUSYBOX_POD -c busybox -- /bin/sh -c "$TEST_CMD"

What's next