Manage clusters from the Google Cloud console

This document explains how to make GKE on Bare Metal available for management in the Google Cloud console. This includes basic management such as being able to log in to clusters and view their workloads, as well as how to enable cluster lifecycle management so that you can upgrade, update, and delete clusters.

Fleet members and the console

All GKE on Bare Metal must be members of a fleet—a unified way to view and manage multiple clusters and their workloads. Each fleet of clusters is associated with a fleet host project.

In GKE on Bare Metal, a user cluster is registered to a fleet at creation time:

  • When you create a cluster using bmctl, you specify your fleet host project in the gkeConnect section of the cluster configuration file. GKE on Bare Metal uses that information to register your cluster to the specified fleet project.

  • When you create a user cluster in the console, the cluster automatically becomes a fleet member in the project selected in the console.

Fleet members outside of Google Cloud like GKE on Bare Metal are displayed in the console in your fleet host project, along with other fleet clusters such as GKE on Google Cloud. The extent to which you can manage GKE on Bare Metal from the console depends on the following:

  • If you have set up authentication, you can sign in to your clusters and view their workloads and other details.

  • If you have enabled cluster lifecycle management for the cluster, you can also upgrade, update, or delete user clusters by using the console. To do this, the cluster must be managed by a service called the GKE On-Prem API. For user clusters created in the console, cluster lifecycle management is enabled at cluster creation time, or you can enable this feature later for user clusters created using bmctl. If this feature is not enabled, you can only manage cluster lifecycle using bmctl on your admin workstation.

View registered clusters

All your fleet clusters are displayed on the Anthos Clusters and the GKE Clusters pages in the console. This both gives you an overview of your entire fleet and, for GKE on Bare Metal, lets you see which clusters are managed by the GKE On-Prem API.

To view your fleet clusters:

  1. In the console, go to the Anthos Clusters page.

    Go to the Anthos Clusters page

  2. Select the Google Cloud project.

    • If Anthos bare metal is displayed in the Type column, the cluster is managed by the GKE On-Prem API.

    • If External is displayed in the Type column, the cluster isn't managed by the GKE On-Prem API.

To see more details about a cluster, yoiu need to sign in and authenticate to the cluster. To do this, you need to do the following steps:

Set up authentication

As described previously, all fleet clusters appear in the GKE and Anthos Clusters lists in the console. However, to see more details such as nodes and workloads (and to perform cluster lifecycle management tasks if the feature is enabled), you need to sign in and authenticate to the cluster. To do this, your registered clusters must be set up with one of the following authentication methods:

  • Google identity: This option lets you sign in using your Google Cloud identity, which is the email address associated with your Google Cloud account. Use this option if users already have access to Google Cloud with their Google identity. If you created the cluster in the console, you can sign in to the cluster using your Google identity, but you will need to configure authentication for other users.

    Login with Google identity is the simplest approach to authentication in the console, so we've described how to set this up in more detail in Set up Google identity authentication.

  • OpenID Connect (OIDC): This option lets you sign in to clusters from the console using their identity from a third-party OIDC identity provider such as Okta or Microsoft AD FS. You might want to use this option if your users have existing usernames, passwords, and security group memberships from your provider. You can find out how to set up third-party OIDC authentication for your clusters in the following guides:

  • Bearer token: If the preceding Google-provided solutions aren't suitable for your organization, you can set up authentication using a Kubernetes service account and using its bearer token to sign in. For details, see Set up using a bearer token.

Grant required roles

Access to the console is controlled by Google Cloud IAM. To manage the cluster lifecycle in the console you need to grant some IAM roles to users who aren't project owners:

  • To let users access the console, at a minimum, you need to grant the following roles:

    • roles/container.viewer. This role lets users view the GKE Clusters page and other container resources in the console. For details about the permissions included in this role, or to grant a role with read and write permissions, see Kubernetes Engine roles in the IAM documentation.

    • roles/gkehub.viewer. This role lets users view clusters outside Google Cloud in the console. For details about the permissions included in this role, or to grant a role with read and write permissions, see GKE Hub roles in the IAM documentation.

  • To let users manage the cluster lifecycle in the console, grant the roles/gkeonprem.admin IAM role. The roles/gkeonprem.admin role gives users administrative access to the GKE On-Prem API, which the console uses to manage the cluster lifecycle. For details on the permissions included in this role, see GKE on-prem roles in the IAM documentation.

The following commands shows how to grant the minimal roles needed to manage cluster lifecycle in the console:

gcloud projects add-iam-policy-binding PROJECT_ID \
    --member=MEMBER \
    --role=roles/container.viewer

gcloud projects add-iam-policy-binding PROJECT_ID \
    --member=MEMBER \
    --role=roles/gkehub.viewer

gcloud projects add-iam-policy-binding PROJECT_ID \
    --member=MEMBER \
    --role=roles/gkeonprem.admin

where:

  • PROJECT_ID is the fleet host project. For clusters created using bmctl, this is the project that you configured in the gkeConnect section of the user cluster's configuration file. For clusters created in the console, this is the project that you chose when the cluster was created.

  • MEMBER is the user's email address in the format user:emailID, for example: user:alice@example.com

Enable cluster lifecycle management in the console

User clusters created in the console are automatically managed by the GKE On-Prem API, and let you perform cluster lifecycle management tasks in the console. If you want to enable this feature for user clusters created using bmctl, follow the steps in Configure a user cluster to be managed by the GKE On-Prem API. When cluster lifecycle management is enabled, you can update clusters, from the console:

  • Update user clusters
  • Add or remove node pools on user clusters
  • Delete user clusters

Set up Google identity authentication

To let users sign in to the cluster using their Google identity, you need to configure the following:

  • Users need specific Identity and Access Management (IAM) roles to be able to see and interact with clusters in the console on the GKE Clusters list and the clusters list pages.

  • Users need to be added to the Kubernetes role-based access control (RBAC) policies that the connect gateway needs to access the cluster Kubernetes API server when using the Connect Agent.

Configure RBAC authorization

Each cluster's Kubernetes API server needs to be able to authorize requests that come from the console. To configure authorization, you need to configure Kubernetes role-based access control (RBAC) policies on each cluster. If you created the cluster in the console, the GKE On-Prem API adds your user account as an administrator and creates appropriate RBAC policies that grant you full administrative access to the cluster.

gcloud CLI

To apply the RBAC policies to users, do the following steps on your admin workstation:

  1. Run the following commands to sign in with your Google Account and update components:

    gcloud auth login
    gcloud components update
    
  2. Generate and apply the RBAC policies to your cluster for users and service accounts:

    gcloud container fleet memberships generate-gateway-rbac  \
        --membership=MEMBERSHIP_NAME \
        --role=ROLE \
        --users=USERS \
        --project=PROJECT_ID \
        --kubeconfig=KUBECONFIG_PATH \
        --context=KUBECONFIG_CONTEXT \
        --apply
    

    Replace the following:

    • MEMBERSHIP_NAME: the name used to uniquely represent the cluster in its fleet. In GKE on Bare Metal, the membership name and the cluster name are the same.
    • ROLE: the Kubernetes role you want to grant to the users on the cluster. To grant users full access to every resource in the cluster in all namespaces, specify clusterrole/cluster-admin. To restrict access, create a custom role, for example: role/mynamespace/namespace-reader. The custom role must already exist before you run the command.
    • USERS: the email addresses of the users (user accounts or service accounts) to whom you want to grant the permissions, as a comma-separated list. For example: --users=foo@example.com,test-acct@test-project.iam.gserviceaccount.com.
    • PROJECT_ID: the project ID of the fleet host project.
    • KUBECONFIG_PATH: the local path of your kubeconfig file that contains an entry for the cluster.
    • KUBECONFIG_CONTEXT: the context of the cluster as it appears in the kubeconfig file. You can get the current context from the command line by running kubectl config current-context. Whether you use the current context or not, make sure that it works for accessing the cluster by running a simple command such as:

      kubectl get namespaces \
        --kubeconfig=KUBECONFIG_PATH \
        --context=KUBECONFIG_CONTEXT
      

    After running gcloud container fleet memberships generate-gateway-rbac, you see something like the following at the end of the output, which is truncated for readability:

    Validating input arguments.
    Specified Cluster Role is: clusterrole/cluster-admin
    Generated RBAC policy is:
    --------------------------------------------
    ...
    Applying the generate RBAC policy to cluster with kubeconfig: /usr/local/google/home/foo/.kube/config, context: kind-kind
    Writing RBAC policy for user: foo@example.com to cluster.
    Successfully applied the RBAC policy to cluster.
    

    This is the context for accessing the cluster through the connect gateway.

    For more details on the generate-gateway-rbac command, see the gcloud CLI reference guide.

bmctl

To apply the RBAC policies to users, do the following steps on your admin workstation:

  1. Add the clusterSecurity.authorization section to your cluster configuration file. Specify your email address and the email address of other users who need to administer the cluster. For example:

    ...
    clusterSecurity:
      authorization:
        clusterAdmin:
          gcpAccounts: [alex@example.com,hao@example.com,sasha@example.com]
    ...
    
  2. Update the cluster:

    bmctl update cluster \
        -c CLUSTER_NAME \
        --kubeconfig=KUBECONFIG
    

    Make the following changes:

    • Replace CLUSTER_NAME with the name of the cluster you want to update.
    • If the cluster is a self-managing cluster (such as admin or standalone cluster), replace KUBECONFIG with the path to the cluster's kubeconfig file. If the cluster is a user cluster, replace KUBECONFIG with the path to the admin cluster's kubeconfig file.

More information