Overview

By using the Gloo Operator to manage your service mesh, you no longer need to manually install and manage the istiod control plane. Instead, you provide minimal Istio configuration to the operator in a ServiceMeshController custom resource, and the operator translates this configuration into a managed istiod control plane in your cluster for you. The operator reduces both the amount of configuration required to deploy Istio, and the overhead required to manage the lifecycle of Istio resources in your cluster.

For more information about the components that are installed in these steps, see the ambient components overview.

Considerations

Before you install an ambient mesh, review the following considerations and requirements.

Version requirements

  • In Gloo Mesh version 2.6 and later, ambient mode requires the Solo distribution of Istio version 1.22.3 or later (1.22.3-solo).
  • If you plan to later link meshes for a multicluster ambient mesh setup: In Gloo Mesh version 2.7 and later, multicluster setups require version 1.24.3 or later.
  • In Istio 1.22.0-1.22.3, the ISTIO_DELTA_XDS environment variable must be set to false. For more information, see this upstream Istio issue. Note that this issue is resolved in Istio 1.22.4.

Single-cluster and multicluster meshes

This guide shows you how to install an ambient mesh in one cluster. If you later decide to to upgrade to a multicluster mesh with an Enterprise level license, you can link your existing ambient meshes across clusters.

Alternatively, if you prefer to start with a multicluster mesh immediately, check out Install and connect new ambient meshes in the multicluster guide instead.

Revision and canary upgrade limitations

The upgrade guides in this documentation show you how to perform in-place upgrades for your Istio components, which is the recommended upgrade strategy.

Set up tools

  1. If you do not already have a license, decide the level of licensed features that you want, and contact an account representative to obtain the license.

  2. Choose the version of Istio that you want to install or upgrade to by reviewing the supported versions table.

  3. Save each value in an environment variable. If you prefer to specify license keys in a secret instead, see Licensing. Note that the Gloo Operator installs the Solo distribution of Istio by default for the version you specify, so neither the -solo image tag nor the repo key are required.

      export GLOO_MESH_LICENSE_KEY=<license_key>
    export ISTIO_VERSION=1.25.2
      
  4. Install or upgrade istioctl with the same version of Istio that you saved.

      curl -L https://istio.io/downloadIstio | ISTIO_VERSION=${ISTIO_VERSION} sh -
    cd istio-${ISTIO_VERSION}
    export PATH=$PWD/bin:$PATH
      

Install the ambient components

  1. Install the Gloo Operator to the gloo-mesh namespace. This operator deploys and manages your Istio installation. Note that if you already installed Gloo Mesh, you can optionally reference the secret that Gloo Mesh automatically creates for your license in the –set manager.env.SOLO_ISTIO_LICENSE_KEY_SECRET_REF=gloo-mesh/license-keys flag instead.

      helm install gloo-operator oci://us-docker.pkg.dev/solo-public/gloo-operator-helm/gloo-operator \
      --version 0.2.3 \
      -n gloo-mesh \
      --create-namespace \
      --set manager.env.SOLO_ISTIO_LICENSE_KEY=${GLOO_MESH_LICENSE_KEY}
      
  2. Verify that the operator pod is running.

      kubectl get pods -n gloo-mesh -l app.kubernetes.io/name=gloo-operator
      

    Example output:

      gloo-operator-78d58d5c7b-lzbr5     1/1     Running   0          48s
      
  3. Create a ServiceMeshController custom resource to configure an Istio installation. For a description of each configurable field, see the ServiceMeshController reference. If you need to set more advanced Istio configuration, you can also create a gloo-extensions-config configmap.
      kubectl apply -n gloo-mesh -f -<<EOF
    apiVersion: operator.gloo.solo.io/v1
    kind: ServiceMeshController
    metadata:
      name: managed-istio
      labels:
        app.kubernetes.io/name: managed-istio
    spec:
      dataplaneMode: Ambient
      installNamespace: istio-system
      version: ${ISTIO_VERSION}
    EOF
      
  4. Verify that the ServiceMeshController is ready. In the Status section of the output, make sure that all statuses are True, and that the phase is SUCCEEDED.

      kubectl describe servicemeshcontroller -n gloo-mesh managed-istio
      

    Example output:

      ...
    Status:
      Conditions:
        Last Transition Time:  2024-12-27T20:47:01Z
        Message:               Manifests initialized
        Observed Generation:   1
        Reason:                ManifestsInitialized
        Status:                True
        Type:                  Initialized
        Last Transition Time:  2024-12-27T20:47:02Z
        Message:               CRDs installed
        Observed Generation:   1
        Reason:                CRDInstalled
        Status:                True
        Type:                  CRDInstalled
        Last Transition Time:  2024-12-27T20:47:02Z
        Message:               Deployment succeeded
        Observed Generation:   1
        Reason:                DeploymentSucceeded
        Status:                True
        Type:                  ControlPlaneDeployed
        Last Transition Time:  2024-12-27T20:47:02Z
        Message:               Deployment succeeded
        Observed Generation:   1
        Reason:                DeploymentSucceeded
        Status:                True
        Type:                  CNIDeployed
        Last Transition Time:  2024-12-27T20:47:02Z
        Message:               Deployment succeeded
        Observed Generation:   1
        Reason:                DeploymentSucceeded
        Status:                True
        Type:                  WebhookDeployed
        Last Transition Time:  2024-12-27T20:47:02Z
        Message:               All conditions are met
        Observed Generation:   1
        Reason:                SystemReady
        Status:                True
        Type:                  Ready
      Phase:                   SUCCEEDED
    Events:                    <none>
      
  5. Verify that the istiod control plane, Istio CNI, and ztunnel pods are running.

      kubectl get pods -n istio-system
      

    Example output:

      NAME                          READY   STATUS    RESTARTS   AGE
    istio-cni-node-6s5nk          1/1     Running   0          2m53s
    istio-cni-node-blpz4          1/1     Running   0          2m53s
    istiod-gloo-bb86b959f-msrg7   1/1     Running   0          2m45s
    istiod-gloo-bb86b959f-w29cm   1/1     Running   0          3m
    ztunnel-mx8nw                 1/1     Running   0          2m52s
    ztunnel-w8r6c                 1/1     Running   0          2m52s
      
  6. Apply the CRDs for the Kubernetes Gateway API to your cluster, which are required to create components such as waypoint proxies for L7 traffic policies, gateways with the Gateway resource, and more.

      kubectl apply -f https://github.com/kubernetes-sigs/gateway-api/releases/download/v1.2.1/standard-install.yaml
      

Next

ServiceMeshController reference

Review the following configurable fields for the ServiceMeshController custom resource.

SettingDescriptionSupported valuesDefault
clusterThe name of the cluster to install Istio into. This value is required to set the trust domain field in multicluster environments.
dataplaneModeThe dataplane mode to use.Ambient or SidecarAmbient
distributionOptional: A specific distribution of the Istio version, such as the standard or FIPS image distribution.Standard or FIPSStandard
image.repositoryOptional: An Istio image repository, such as to use an image from a private registry.The Solo distribution of Istio repo for the Istio minor version.
image.secretsOptional: A list of secrets to use for pulling images from a container registry. The secret list must be of type kubernetes.io/dockerconfigjson and exist in the installNamespace that you install Istio in.
installNamespaceNamespace to install the service mesh components into. If you set the installNamespace to a namespace other than gloo-system, gloo-mesh, or istio-system, you must include the –set manager.env.WATCH_NAMESPACES=<namespace> setting.istio-system
networkThe default network where workload endpoints exist. A network is a logical grouping of workloads that exist in the same Layer 3 domain. Workloads in the same network can directly communicate with each other, while workloads in different networks require an east-west gateway to establish connectivity. This value is required in multi-network environments. For example, an easy way to identify the network of in-mesh workloads in one cluster is to simply use the cluster’s name for the network, such as cluster1.
onConflictOptional: How to resolve conflicting Istio configuration, if the configuration in this ServiceMeshController conflicts with existing Istio resources in the cluster.
  • Force: The existing resources are updated with the new configuration.
  • Abort: The installation configured in this ServiceMeshController is aborted, and the existing resources remain unchanged.
Force or AbortAbort
repository.secretsOptional: A list of secrets to use for pulling manifests from an artifact registry. The secret list must be of type kubernetes.io/dockerconfigjson and can exist in any namespace, such as the same namespace that you create the ServiceMeshController in.
repository.insecureSkipVerifyOptional: If set to true, the repository server’s certificate chain and hostname are not verified.true or false
scalingProfileOptional: The istiod control plane scaling settings to use. In large environments, set to Large.
  • Default sets the following scaling values:
    • resources.requests.cpu=1000m
    • resources.requests.memory=1Gi
    • resources.limits.cpu=2000m
    • resources.limits.memory=2Gi
    • autoscaleEnabled=true
    • autoscaleMin=2
    • autoscaleMax=25
    • cpu.targetAverageUtilization=80
  • Demo sets the following scaling values:
    • autoscaleEnabled=false
    • resources.requests.cpu=250m
  • Large sets the following scaling values:
    • resources.requests.cpu=4000m
    • resources.requests.memory=4Gi
    • resources.limits.cpu=4000m
    • resources.limits.memory=4Gi
    • autoscaleEnabled=true
    • autoscaleMin=4
    • autoscaleMax=50
    • cpu.targetAverageUtilization=75
Default, Demo, or LargeDefault
trafficCaptureModeOptional: Traffic capture mode to use.
  • Auto: The most suitable traffic capture mode is automatically selected based on the environment, such as using a CNI to capture traffic.
  • InitContainer: An init container is used for the traffic capture. This setting can only be used when the dataplaneMode is Sidecar.
Auto or InitContainerAuto
trustDomainThe trustDomain for Istio workloads.If cluster is set, defaults to that value. If cluster is unset, defaults to cluster.local.
versionThe Istio patch version to install. For more information, see Supported Solo distributions of Istio.Any Istio version supported for your Gloo version

Advanced settings configuration

You can set advanced Istio configuation by creating a configmap. For example, you might need to specify settings for istiod such as discovery selectors, pod and service annotations, affinities, tolerations, or node selectors.

The following gloo-extensions-config example configmap sets all possible fields for demonstration purposes.

  apiVersion: v1
kind: ConfigMap
metadata:
  name: gloo-extensions-config
  namespace: gloo-mesh
data:
  stable: |
    serviceMeshController:
      istiod:
        discoverySelectors:
          - matchLabels:
              foo: bar
        topology:
          affinity:
            podAntiAffinity:
              preferredDuringSchedulingIgnoredDuringExecution:
              - podAffinityTerm:
                  labelSelector:
                    matchExpressions:
                    - key: foo
                      operator: In
                      values:
                      - bar
                  topologyKey: foo.io/bar
                weight: 80
          nodeSelector:
            foo: bar
          tolerations:
            - key: t1
              operator: Equal
              value: v1
        deployment:
          podAnnotations:
            foo: bar
        serviceAnnotations:
          foo: bar
  beta: |
    serviceMeshController:
      cni:
        confDir: /foo/bar
        binDir: /foo/bar