Istio

Limited vCluster Tenancy Configuration Support

This feature is only available for the following:

Running the control plane as a container and the following worker node types:

Host Nodes

Enterprise-Only Feature

This feature is an Enterprise feature. See our pricing plans or contact our sales team for more information.

Istio integration

This guide shows how to set up Istio integration with your virtual cluster. This enables you to use one Istio installation from the host cluster instead of installing Istio in each virtual cluster.

You can include your virtual workloads in the mesh by setting istio.io/dataplane-mode=ambient label on the virtual Namespaces or Pods. You can exclude your virtual workloads from the mesh by setting istio.io/dataplane-mode=none label either on the Namespace or on the Pod.

Prerequisites

• Administrator access to a Kubernetes cluster: See Accessing Clusters with kubectl for more information. Run the command kubectl auth can-i create clusterrole -A to verify that your current kube-context has administrative privileges.

  info

  To obtain a kube-context with admin access, ensure you have the necessary credentials and permissions for your Kubernetes cluster. This typically involves using kubectl config commands or authenticating through your cloud provider's CLI tools.

• helm: Helm v3.10 is required for deploying the platform. Refer to the Helm Installation Guide if you need to install it.

• kubectl: Kubernetes command-line tool for interacting with the cluster. See Install and Set Up kubectl for installation instructions.

• istio operator installed on your host cluster in ambient mode with DNS Capture disabled

warning

To disable DNS capture, set values.cni.ambient.dnsCapture: false in your Istio configuration. This integration works only with Istio in Ambient mode. Sidecar mode is not supported.

Enable the integration

Enable the Istio integration in your virtual cluster configuration:

Enable istio integration

integrations:
  istio:
    enabled: true

This configuration:

• Enables the integration.
• Installs Resource Definitions for DestinationRules, Gateways and VirtualServices into the virtual cluster.
• Exports DestinationRules, Gateways and VirtualServices from the virtual cluster to the host (and re-writes) service references to the services translated names in the host.
• Adds istio.io/dataplane-mode label to the synced Pods based on the value of this label set in the virtual namespace.

warning

Only DestinationRules, Gateways, and VirtualServices from networking.istio.io/v1 API Version are synced to the host clusters. Other kinds are not yet supported.

Set up cluster contexts

Setting up the host and virtual cluster contexts makes it easier to switch between them.

export HOST_CTX="your-host-context"
export VCLUSTER_CTX="vcluster-ctx"
export VCLUSTER_HOST_NAMESPACE="vcluster"

tip

You can find your contexts by running kubectl config get-contexts

Route request based on the version label of the app

Create waypoint proxy in the host

1. In this tutorial, you set Kubernetes service name as a host in the VirtualService spec.hosts. To make it work, you need a Waypoint proxy in the virtual cluster's host namespace. In many cases it is optional however. Refer to Istio documentation for more information on Waypoint proxies. Install Gateway CRD first in the host:

   Install Gateway CRD

   kubectl --context="${HOST_CTX}" get crd gateways.gateway.networking.k8s.io &> /dev/null || \
   kubectl --context="${HOST_CTX}" apply -f https://github.com/kubernetes-sigs/gateway-api/releases/download/v1.2.1/standard-install.yaml

   this is a Gateway for Waypoint you need:

   waypoint-gateway.yaml

   apiVersion: gateway.networking.k8s.io/v1
   kind: Gateway
   metadata:
     name: waypoint
     labels:
       istio.io/waypoint-for: service
   spec:
     gatewayClassName: istio-waypoint
     listeners:
     - name: mesh
       port: 15008
       protocol: HBONE

   create it in the host cluster:

   Create Waypoint Gateway

   kubectl --context="${HOST_CTX}" create -f waypoint-gateway.yaml --namespace="${VCLUSTER_HOST_NAMESPACE}"

Create virtual namespace with ambient mode enabled

2. First, you create test namespace:

   Create test namespace

   kubectl --context="${VCLUSTER_CTX}" create namespace test

   and label it with istio.io/dataplane-mode: ambient:

   Label test namespace

   kubectl --context="${VCLUSTER_CTX}" label namespace test istio.io/dataplane-mode=ambient

Create two versions of your app

3. Next, you create 3 deployments: two of them are nginx server and the third one is to curl the other two.

   Create NGINX deployments that respond with different response bodies based on the contents of their respective ConfigMaps:

   configmap1.yaml

   apiVersion: v1
   kind: ConfigMap
   metadata:
     name: nginx-configmap-v1
     namespace: test
   data:
     index.html: |
       <!DOCTYPE html>
       <html>
       <head>
         <title>Welcome to nginx v1!</title>
       </head>
       <body>
       <h1>Hello from Nginx Version 1!</h1>
       </body>
       </html>

   deployment1.yaml

   apiVersion: apps/v1
   kind: Deployment
   metadata:
     name: nginx-deployment-v1
     namespace: test
     labels:
       app: nginx
       version: v1
   spec:
     replicas: 1
     selector:
       matchLabels:
         app: nginx
         version: v1
     template:
       metadata:
         labels:
           app: nginx
           version: v1
       spec:
         containers:
         - name: nginx
           image: nginx:latest
           ports:
           - containerPort: 80
           volumeMounts:
           - name: nginx-index-v1
             mountPath: /usr/share/nginx/html/index.html
             subPath: index.html
         volumes:
         - name: nginx-index-v1
           configMap:
             name: nginx-configmap-v1

   Create v1 config map

   kubectl --context="${VCLUSTER_CTX}" create -f configmap1.yaml --namespace test

   Create v1 deployment

   kubectl --context="${VCLUSTER_CTX}" create -f deployment1.yaml --namespace test

   make sure that this nginx app is up and running:

   Wait for v1 pods

   kubectl --context="${VCLUSTER_CTX}" wait --for=condition=ready pod -l app=nginx --namespace test --timeout=300s

   Create an additional NGINX deployment configured to serve a different response body, using a separate ConfigMap:

   configmap2.yaml

     apiVersion: v1
     kind: ConfigMap
     metadata:
       name: nginx-configmap-v2
       namespace: test
     data:
       index.html: |
         <!DOCTYPE html>
         <html>
         <head>
           <title>Welcome to nginx v2!</title>
         </head>
         <body>
           <h1>Hello from Nginx Version 2!</h1>
         </body>
   </html>

   deployment2.yaml

   apiVersion: apps/v1
   kind: Deployment
   metadata:
     name: nginx-deployment-v2
     namespace: test
     labels:
       app: nginx
       version: v2
   spec:
     replicas: 1
     selector:
       matchLabels:
         app: nginx
         version: v2
     template:
       metadata:
         labels:
           app: nginx
           version: v2
       spec:
         containers:
           - name: nginx
             image: nginx:latest
             ports:
               - containerPort: 80
             volumeMounts:
               - name: nginx-index-v2
                 mountPath: /usr/share/nginx/html/index.html
                 subPath: index.html
         volumes:
           - name: nginx-index-v2
             configMap:
               name: nginx-configmap-v2

   Create v2 config map

   kubectl --context="${VCLUSTER_CTX}" create -f configmap2.yaml --namespace test

   Create v2 deployment

   kubectl --context="${VCLUSTER_CTX}" create -f deployment2.yaml --namespace test

   To ensure your NGINX application is up and running in your Kubernetes cluster, use the following command:

   Wait for v2 pods

   kubectl --context="${VCLUSTER_CTX}" wait --for=condition=ready pod -l app=nginx --namespace test --timeout=300s

   Create a Service that targets Pods from both Deployments by using a shared label:

   service.yaml

   apiVersion: v1
   kind: Service
   metadata:
     name: nginx-service
     namespace: test
     labels:
       app: nginx
       istio.io/use-waypoint: "waypoint"
   spec:
     ports:
       - port: 80
         targetPort: 80
     selector:
       app: nginx

   The istio.io/use-waypoint: waypoint label directs Istio to route traffic for the labeled resource through the waypoint proxy within the same namespace. This configuration enables Layer 7 (L7) policy enforcement and observability features provided by the waypoint proxy. Applying this label to a namespace ensures that all Pods and Services within that namespace use the specified waypoint proxy.

   To deploy a Service defined in the service.yaml file within the test namespace of the Kubernetes cluster specified by the ${VCLUSTER_CTX} context, use the following command:

   Create service

   kubectl --context="${VCLUSTER_CTX}" create -f service.yaml --namespace test

   To test connectivity between the two NGINX deployments, deploy a temporary Pod equipped with curl:

   client_deployment.yaml

   apiVersion: apps/v1
   kind: Deployment
   metadata:
     name: client
     namespace: test
     labels:
       app: client
   spec:
     replicas: 1
     selector:
       matchLabels:
         app: client
     template:
       metadata:
         labels:
           app: client
       spec:
         containers:
           - name: nginx
             image: nginx:latest
             ports:
               - containerPort: 80

   Create client deployment

   kubectl --context="${VCLUSTER_CTX}" create -f client_deployment.yaml --namespace test

Configure your desired traffic routing using DestinationRule and VirtualService

4. You can create DestinationRules and VirtualService in the virtual cluster.

   Create a pair that routes our request based on the request path:

   1. Requesting /v2 endpoint should route our request to pods with version=v2 label
   2. All other requests are routed to version=v1 pods.

   Save this DestinationRule and VirtualService definition, and apply it in the virtual cluster:

   destination_rule.yaml

   apiVersion: networking.istio.io/v1
   kind: DestinationRule
   metadata:
     name: nginx-destination
     namespace: test
   spec:
     host: nginx-service.test.svc.cluster.local # vCluster translates it to the host service automatically
     subsets:
       - name: v1
         labels:
           version: v1
       - name: v2
         labels:
           version: v2

   virtual_service.yaml

   apiVersion: networking.istio.io/v1
   kind: VirtualService
   metadata:
     name: nginx-service
     namespace: test
   spec:
     hosts:
       - nginx-service.test.svc.cluster.local # vCluster translates it to the host service automatically
     http:
       - name: "nginx-v2"
         match:
           - uri:
               prefix: "/v2"
         rewrite:
           uri: "/"
         route:
           - destination:
               host: nginx-service.test.svc.cluster.local # vCluster translates it to the host service automatically
               subset: v2
       - name: "nginx-v1"
         route:
           - destination:
               host: nginx-service.test.svc.cluster.local # vCluster translates it to the host service automatically
               subset: v1
   

   To apply a DestinationRule configuration to the virtual cluster specified by the ${VCLUSTER_CTX} context, use the following command:

   Create destination rule

   kubectl --context="${VCLUSTER_CTX}" create -f destination_rule.yaml

   Create virtual service

   kubectl --context="${VCLUSTER_CTX}" create -f virtual_service.yaml

Verify that DestinationRule and VirtualService is synced to the host cluster

5. Check destination rule in the host cluster

   kubectl --context="${HOST_CTX}" get destinationrules --namespace "${VCLUSTER_HOST_NAMESPACE}"

   Check virtual service in the host cluster

   kubectl --context="${HOST_CTX}" get virtualservices --namespace "${VCLUSTER_HOST_NAMESPACE}"

   You should see a DestinationRule named nginx-destination-x-test-x-vcluster and VirtualService named nginx-service-x-test-x-vcluster.

Test traffic routing

6. Execute a curl command from within the client Pod to verify responses from the two NGINX deployments. Depending on the request path, you should receive either "Hello from Nginx Version 1!" or "Hello from Nginx Version 2!" in the response:

   Query version 2

   kubectl --context="${VCLUSTER_CTX}" exec -it -n test deploy/client -- curl nginx-service/v2
   <!DOCTYPE html>
   <html>
   <head>
     <title>Welcome to nginx v2!</title>
   </head>
   <body>
   <h1>Hello from Nginx Version 2!</h1>
   </body>
   </html>

   Query version 1

   kubectl --context="${VCLUSTER_CTX}" exec -it -n test deploy/client -- curl nginx-service
   <!DOCTYPE html>
   <html>
   <head>
     <title>Welcome to nginx v1!</title>
   </head>
   <body>
   <h1>Hello from Nginx Version 1!</h1>
   </body>
   </html>

   Seeing the same output means that request was intercepted by Istio and routed as we specified in the DestinationRule and VirtualService.

Summary

7. Istio integration enables you to re-use one Istio instance from the host cluster for multiple virtual clusters. Virtual cluster users can define their own Gateway, DestinationRule and VirtualService without interfering with each other.

Fields translated during the sync to host

Following fields of Gateway are modified by vCluster during the sync to host:

• reference to the TLS Secret is re-written spec.servers[*].tls.credentialName. Secret is automatically synced to the host cluster.
• namespace, . and * prefix, followed by / is stripped from spec.servers[*].hosts[*], so e.g. foo-namespace/loft.sh becomes loft.sh in the host object.
• additional labels vcluster.loft.sh/managed-by: [YOUR VIRTUAL CLUSTER NAME] and vcluster.loft.sh/namespace: [VIRTUAL NAMESPACE] are automatically added to the spec.subsets[*].labels

For additional information how Secret and Service references are translated, read How does syncing work?

Following fields of DestinationRule are modified by vCluster during the sync to host:

• reference to the virtual Kubernetes Service is re-written for spec.host
• reference to the TLS Secret in spec.trafficPolicy.portLevelSettings[*].tls.credentialName & spec.trafficPolicy.tls.credentialName is re-written. Secrets are automatically synced to the host cluster.
• additional labels

Following fields of VirtualService are modified by vCluster during the sync to host:

• reference to the virtual Kubernetes Service is re-written for:

• spec.hosts[*]

• spec.http[*].route[*].destination.host

• spec.http[*].mirrors[*].destination.host

• spec.tcp[*].route[*].destination.host

• spec.tls[*].route[*].destination.host

  • reference to the networking.istio.io/v1 kind: Gateway is re-written for:

• spec.gateways[*]

• spec.http[*].match[*].gateways[*]

• spec.tls[*].match[*].gateways[*]

• spec.tcp[*].match[*].gateways[*]

  • reference to the networking.istio.io/v1 kind: VirtualService is re-written for:

• spec.http[*].delegate

Fields not supported in VirtualService:

• spec.exportTo
• spec.http[*].match[*].sourceLabels
• spec.http[*].match[*].sourceNamespace
• spec.tcp[*].match[*].sourceLabels
• spec.tcp[*].match[*].sourceNamespace
• spec.tls[*].match[*].sourceLabels
• spec.tls[*].match[*].sourceNamespace

Config reference

istio required object

Istio syncs DestinationRules, Gateways and VirtualServices from virtual cluster to the host.

enabled required boolean false

Enabled defines if this option should be enabled.

sync required object

toHost required object

destinationRules required object

enabled required boolean true

Enabled defines if this option should be enabled.

gateways required object

enabled required boolean true

Enabled defines if this option should be enabled.

virtualServices required object

enabled required boolean true

Enabled defines if this option should be enabled.