1. Documentation
×

Installing a cluster on Google Cloud Dedicated

Prerequisites for installing a cluster on Google Cloud Dedicated

Installing OKD on GCD requires a configured project with domain-scoped IDs, Workforce Identity Federation credentials, and network access to sovereign cloud API endpoints.

Google Cloud Dedicated overview

Google Cloud Dedicated (GCD) is a sovereign cloud platform that provides Google Cloud technology in a fully isolated environment with strict data and operational sovereignty guarantees.

OKD supports deploying clusters to GCD sovereign cloud regions by using installer-provisioned infrastructure.

You cannot select a GCD sovereign cloud region by using the guided terminal prompts from the installation program. You must define the region and other GCD-specific parameters manually in the install-config.yaml file.

Sovereign cloud framework

GCD is part of Google’s sovereign cloud strategy, which addresses data residency, administrative access control, and operational sovereignty. Google’s sovereign cloud portfolio includes the following tiers:

  • Assured Workloads provides data residency controls within the public cloud.

  • Sovereign Controls by Partners adds partner-managed key management and audits.

  • GCD provides the highest level of sovereignty with fully isolated infrastructure operated by a regional partner.

Each GCD deployment is operated by a local partner who controls administrative access to the infrastructure. For example, in Germany, the main GCD partner is Thales.

Key differences from public Google Cloud

GCD regions differ from public Google Cloud in the following ways that affect OKD installation and operation:

Single region per universe

Each GCD deployment operates as a single region. Multi-region features such as cross-region load balancing and multi-region storage are not available. You must use multiple zones within the single region for high availability.

Different API endpoints

GCD uses API endpoints in the format <service>.apis-<region-host>.goog instead of <service>.googleapis.com. For example, the Berlin region uses compute.apis-berlin-build0.goog instead of compute.googleapis.com. Cluster Operators automatically detect the sovereign cloud environment and override standard googleapis.com endpoints.

Domain-scoped project IDs

All GCD project IDs carry a mandatory prefix, such as eu0:. For example, a project named my-project has the project ID eu0:my-project. You must include this prefix in all commands and configuration.

Different service account email format

Service account email addresses use GCD-specific domains instead of standard Google Cloud format.

Limited service availability

Only a subset of Google Cloud services is available. Older Compute Engine machine types, ARM-based instances, TPUs, and several other services are not available.

No default VPC

A default VPC network is not automatically created for new projects. You must create or configure a VPC explicitly.

OKD constraints on GCD

OKD on GCD has constraints around machine types, disk types, operating system images, DNS, and endpoint configuration that differ from standard Google Cloud installations.

Supported GCD regions

The following GCD sovereign cloud regions are supported for OKD installation:

Table 1. GCD regions
Region Identifier Zones

Berlin, Germany

u-germany-northeast1

u-germany-northeast1-a, u-germany-northeast1-b, u-germany-northeast1-c

Additional GCD regions might be supported in future releases. The architectural constraints, such as the single-region design, limited machine types, and different API endpoints, apply to all GCD regions.

Limitations and considerations for Google Cloud Dedicated

GCD environments have specific constraints for machine types, storage, networking, and identity that differ from standard Google Cloud installations to maintain sovereignty and data isolation requirements.

Machine types and disk types

GCD has the following limitations for machine types and disk types compared to public Google Cloud:

  • Only the C3, M3, and A3 Edge machine series are available. The installation program defaults to c3-standard-4 for all machine pools. If you specify a machine type from an unavailable series, the installation fails.

  • Only the hyperdisk-balanced disk type is available. The installation program defaults to hyperdisk-balanced for sovereign cloud installations. Other disk types such as pd-ssd, pd-standard, and Local SSD are not available.

  • The installation program validates disk type availability against the GCD API on a per-zone basis. If a disk type is not available in the specified zones, the installation fails with an error indicating the unavailable zones.

  • ARM-based (aarch64) machine types and images are not available.

Storage

GCD has the following storage limitations compared to public Google Cloud:

  • You must change the default storage class to use hyperdisk-balanced after installation. The default storage class provisioned by the installation program uses pd-ssd, which is not available in GCD regions. For more information, see "Changing the default storage class".

  • Cloud Storage is available but limited to single-region buckets only. Dual-region, multi-region, and bucket relocation are not available.

  • Cloud Storage bucket locations must be set explicitly. There is no default bucket location in GCD.

  • Google Cloud Filestore is not available.

Operating system image

The Fedora CoreOS (FCOS) image is not pre-published in GCD regions. You must provide a custom operating system image that you uploaded to your GCD project.

In the install-config.yaml file, specify the operating system image by using the platform.gcp.defaultMachinePlatform.osImage field or the per-machine-pool fields for osImage. You must provide both the name and project fields. If these fields are not set, the installation fails.

Example custom operating system image configuration
platform:
  gcp:
    defaultMachinePlatform:
      osImage:
        name: my-custom-rhcos
        project: eu0:my-image-project

Networking

GCD has the following networking restrictions:

  • Only private DNS zones are supported. Public DNS zones are not available in GCD. You must set the publish parameter to Internal in the install-config.yaml file.

  • Private Service Connect (PSC) endpoint overrides are not supported. Do not set the platform.gcp.endpoint field in the install-config.yaml file. PSC endpoint overrides generate URLs targeting the googleapis.com domain, which is not compatible with GCD sovereign cloud environments. Features that depend on PSC endpoints, such as Hive, are not supported.

  • A default VPC network is not automatically created for new GCD projects. You must create or specify a VPC network before installation.

Identity and service accounts

GCD has the following identity and service account limitations compared to public Google Cloud:

  • Email addresses for service accounts use a different domain format in GCD. For domain-scoped project IDs, such as eu0:my-project, the installation program generates service account emails in the format <name>@<project-name>.<prefix>.iam.gserviceaccount.com instead of the standard <name>@<project-id>.iam.gserviceaccount.com.

  • Google Accounts and Google Groups are not supported for IAM bindings. Use a service account.

  • The GCD credential file does not populate the project_id field. The installation program reads the project ID from the platform.gcp.projectID field in the install-config.yaml file instead.

Encryption and Key Management Service

GCD has the following encryption and Key Management Service requirements:

  • When you specify the defaultMachinePlatform field in the install-config.yaml file, the installation program encrypts both the image registry and the bootstrap ignition by using Key Management Service (KMS). Global KMS keys are not accepted. You must use regional KMS keys that reside in the same GCD region as your cluster.

  • You must update Identity and Access Management (IAM) policies to grant the required permissions for your KMS keys before installation.

Other limitations

GCD has the following additional limitations:

  • The installation program automatically detects a GCD sovereign cloud environment based on the domain-scoped project ID prefix (such as eu0:) and the region prefix (such as u-).

  • The universe domain from the GCD credentials is recorded on the cluster’s Infrastructure custom resource at status.platformStatus.gcp.universeDomain. Cluster components use this value to route API requests to the correct GCD endpoints.

  • Multi-project installations that use Shared VPC (XPN) are not supported.

Additional resources

Preparing your Google Cloud Dedicated environment

To ensure your OKD cluster can authenticate and communicate with GCD’s isolated infrastructure, you configure project settings, credentials, and networking that comply with sovereignty requirements.

In addition to the standard Google Cloud project configuration, GCD requires configurations described in the following procedure.

Prerequisites

  • You have a GCD project with a domain-scoped project ID, such as eu0:my-project.

  • You have a key file for a service account that contains the universe domain.

  • You have a service account with the required permissions for OKD installation.

Procedure

  1. Confirm the following environment details with your GCD operating partner, such as Thales:

    • The region identifier for your GCD deployment, such as u-germany-northeast1.

    • The universe domain for your region, such as apis-berlin-build0.goog.

    • Your domain-scoped project ID, such as eu0:my-project.

    • Network connectivity requirements for your sovereign cloud region.

  2. Obtain the GCD credential configuration file for your service account. The credential file must include the universe_domain field that corresponds to your GCD region.

    Verify that your credential file contains the universe_domain field by running the following command:

    $ cat <path_to_credentials>/credentials.json | grep universe_domain
    Example output
      "universe_domain": "apis-berlin-build0.goog"

  3. Set the GOOGLE_CLOUD_UNIVERSE_DOMAIN environment variable to the universe domain for your GCD region by running the following command:

    $ export GOOGLE_CLOUD_UNIVERSE_DOMAIN=<universe_domain>

    where:

    <universe_domain>

    Specifies the universe domain for your GCD region. For the Berlin region, use apis-berlin-build0.goog.

  4. Set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the path of your GCD credential file by running the following command:

    $ export GOOGLE_APPLICATION_CREDENTIALS=<path_to_credentials>/credentials.json

    where:

    <path_to_credentials>

    Specifies the path to the directory that contains your GCD credential file.

  5. Verify that the required APIs are enabled in your GCD project:

    • Compute Engine API (compute.googleapis.com)

    • Cloud DNS API (dns.googleapis.com)

    • Cloud Storage API (storage.googleapis.com)

    • IAM API (iam.googleapis.com)

    • Service Usage API (serviceusage.googleapis.com)

    • Cloud Resource Manager API (cloudresourcemanager.googleapis.com)

    GCD uses the same API service names as public Google Cloud for enabling and disabling APIs. The service endpoints, however, use the GCD-specific domain.

  6. Create a VPC network if one does not already exist by running the following command. GCD does not create a default VPC network for new projects.

    $ gcloud compute networks create <network_name> \
    --subnet-mode=custom \
    --project=<gcd_project_id>

    where:

    <network_name>

    Specifies the name of the VPC network to create.

    <gcd_project_id>

    Specifies your domain-scoped GCD project ID, such as eu0:my-project.

  7. Configure a private DNS zone for your cluster’s base domain by running the following command. GCD supports only private DNS zones.

    $ gcloud dns managed-zones create <zone_name> \
    --dns-name=<base_domain> \
    --description="Private DNS zone for OpenShift" \
    --visibility=private \
    --networks=<network_name> \
    --project=<gcd_project_id>

    where:

    <zone_name>

    Specifies a name for the DNS zone.

    <base_domain>

    Specifies the base domain for your cluster.

    <network_name>

    Specifies the name of the VPC network you created.

    <gcd_project_id>

    Specifies your domain-scoped GCD project ID, such as eu0:my-project.

  8. Prepare a custom Fedora CoreOS (FCOS) image for your GCD project. The FCOS image is not pre-published in GCD regions. You must upload your own image:

    1. Download the FCOS image for your OKD version from the FCOS image mirror.

      Download the gcp.x86_64.tar.gz file for your version.

    2. Create a Cloud Storage bucket in your GCD project to store the image by running the following command:

      $ gcloud storage buckets create gs://<bucket_name> \
    --location=<gcd_region> \
    --project=<gcd_project_id>

      where:

      <bucket_name>

      Specifies a unique name for the storage bucket.

      <gcd_region>

      Specifies your GCD region name, such as u-germany-northeast1.

      <gcd_project_id>

      Specifies your domain-scoped GCD project ID, such as eu0:my-project.

      Cloud Storage in GCD is limited to single-region buckets. The bucket location must match your cluster region.

    3. Upload the FCOS image to your Cloud Storage bucket by running the following command:

      $ gcloud storage cp rhcos-<version>-gcp.x86_64.tar.gz \
    gs://<bucket_name>/ \
    --project=<gcd_project_id>

      Replace <version> with the version you downloaded.

    4. Import the FCOS image from Cloud Storage to create a Compute Engine image by running the following command:

      $ gcloud compute images create <image_name> \
    --source-uri=gs://<bucket_name>/rhcos-<version>-gcp.x86_64.tar.gz \
    --guest-os-features=GVNIC,UEFI_COMPATIBLE,VIRTIO_SCSI_MULTIQUEUE \
    --project=<gcd_project_id>

      where:

      <image_name>

      Specifies a name for the FCOS image.

      --source-uri

      Specifies the Cloud Storage URI (gs://) where the image file was uploaded in the previous step. The gs:// URI must point to the Cloud Storage bucket in your GCD project.

      --guest-os-features

      Specifies the guest operating system features required for FCOS images. You must enable the GVNIC, UEFI_COMPATIBLE, and VIRTIO_SCSI_MULTIQUEUE features for proper network, UEFI boot support, and disk I/O performance.

      <bucket_name>

      Specifies the name of the Cloud Storage bucket you created.

      <version>

      Specifies the FCOS version you downloaded.

      <gcd_project_id>

      Specifies your domain-scoped GCD project ID, such as eu0:my-project.

    5. Note the image name and project ID. You must specify these values in the platform.gcp.defaultMachinePlatform.osImage fields of the install-config.yaml file.

Generating a key pair for cluster node SSH access

During an OKD installation, you can provide an SSH public key to the installation program. The key is passed to the Fedora CoreOS (FCOS) nodes through their Ignition config files and is used to authenticate SSH access to the nodes. The key is added to the ~/.ssh/authorized_keys list for the core user on each node, which enables password-less authentication.

The key is added to the ~/.ssh/authorized_keys list for the core user on each node, which enables password-less authentication. After the key is passed to the nodes, you can use the key pair to SSH in to the FCOS nodes as the user core. To access the nodes through SSH, the private key identity must be managed by SSH for your local user.

If you want to SSH in to your cluster nodes to perform installation debugging or disaster recovery, you must provide the SSH public key during the installation process. The ./openshift-install gather command also requires the SSH public key to be in place on the cluster nodes.

Do not skip this procedure in production environments, where disaster recovery and debugging is required.

You must use a local key, not one that you configured with platform-specific approaches.

On clusters running Fedora CoreOS (FCOS), the SSH keys specified in the Ignition config files are written to the /home/core/.ssh/authorized_keys.d/core file. However, the Machine Config Operator manages SSH keys in the /home/core/.ssh/authorized_keys file and configures sshd to ignore the /home/core/.ssh/authorized_keys.d/core file. As a result, newly provisioned OKD nodes are not accessible using SSH until the Machine Config Operator reconciles the machine configs with the authorized_keys file. After you can access the nodes using SSH, you can delete the /home/core/.ssh/authorized_keys.d/core file.
Procedure

  1. If you do not have an existing SSH key pair on your local machine to use for authentication onto your cluster nodes, create one. For example, on a computer that uses a Linux operating system, run the following command:

    $ ssh-keygen -t ed25519 -N '' -f <path>/<file_name>

    Specifies the path and file name, such as ~/.ssh/id_ed25519, of the new SSH key. If you have an existing key pair, ensure your public key is in the your ~/.ssh directory.

    If you plan to install an OKD cluster that uses the Fedora cryptographic libraries that have been submitted to NIST for FIPS 140-2/140-3 Validation on only the x86_64, ppc64le, and s390x architectures, do not create a key that uses the ed25519 algorithm. Instead, create a key that uses the rsa or ecdsa algorithm.

  2. View the public SSH key:

    $ cat <path>/<file_name>.pub

    For example, run the following to view the ~/.ssh/id_ed25519.pub public key:

    $ cat ~/.ssh/id_ed25519.pub

  3. Add the SSH private key identity to the SSH agent for your local user, if it has not already been added. SSH agent management of the key is required for password-less SSH authentication onto your cluster nodes, or if you want to use the ./openshift-install gather command.

    On some distributions, default SSH private key identities such as ~/.ssh/id_rsa and ~/.ssh/id_dsa are managed automatically.

    1. If the ssh-agent process is not already running for your local user, start it as a background task:

      $ eval "$(ssh-agent -s)"
      Example output
      Agent pid 31874

      If your cluster is in FIPS mode, only use FIPS-compliant algorithms to generate the SSH key. The key must be either RSA or ECDSA.

  4. Add your SSH private key to the ssh-agent:

    $ ssh-add <path>/<file_name>

    Specify the path and file name for your SSH private key, such as ~/.ssh/id_ed25519.

    Example output
    Identity added: /home/<you>/<path>/<file_name> (<computer_name>)
Next steps

  • When you install OKD, provide the SSH public key to the installation program.

Obtaining the installation program

Before you install OKD, download the installation file on the host you are using for installation, so that installation assets exist for deployment in your environment.

Prerequisites

  • You have a computer that runs Linux or macOS, with 500 MB of local disk space.

Procedure

  1. Download the installation program from https://github.com/openshift/okd/releases.

    • The installation program creates several files on the computer that you use to install your cluster. You must keep the installation program and the files that the installation program creates after you finish installing the cluster. Both of the files are required to delete the cluster.

    • Deleting the files created by the installation program does not remove your cluster, even if the cluster failed during installation. To remove your cluster, complete the OKD uninstallation procedures for your specific cloud provider.

  2. Extract the installation program. For example, on a computer that uses a Linux operating system, run the following command:

    $ tar -xvf openshift-install-linux.tar.gz

  3. Download your installation pull secret from Red Hat OpenShift Cluster Manager. This pull secret allows you to authenticate with the services that are provided by the included authorities, including Quay.io, which serves the container images for OKD components.

    Using a pull secret from Red Hat OpenShift Cluster Manager is not required. You can use a pull secret for another private registry. Or, if you do not need the cluster to pull images from a private registry, you can use {"auths":{"fake":{"auth":"aWQ6cGFzcwo="}}} as the pull secret when prompted during the installation.

    If you do not use the pull secret from Red Hat OpenShift Cluster Manager:

    • Red Hat Operators are not available.

    • The Telemetry and Insights Operators do not send data to Red Hat.

    • Content from the Red Hat Ecosystem Catalog Container images registry, such as image streams and Operators, are not available.

Manually creating the installation configuration file

Installing the cluster requires that you manually create the installation configuration file.

Prerequisites

  • You have an SSH public key on your local machine for use with the installation program. You can use the key for SSH authentication onto your cluster nodes for debugging and disaster recovery.

  • You have obtained the OKD installation program and the pull secret for your cluster.

Procedure

  1. Create an installation directory to store your required installation assets in:

    $ mkdir <installation_directory>

    You must create a directory. Some installation assets, such as bootstrap X.509 certificates have short expiration intervals, so you must not reuse an installation directory. If you want to reuse individual files from another cluster installation, you can copy them into your directory. However, the file names for the installation assets might change between releases. Use caution when copying installation files from an earlier OKD version.

  2. Customize the provided sample install-config.yaml file template and save the file in the <installation_directory>.

    You must name this configuration file install-config.yaml.

  3. Back up the install-config.yaml file so that you can use it to install many clusters.

    Back up the install-config.yaml file now, because the installation process consumes the file in the next step.
Additional resources

Sample install-config.yaml file for Google Cloud Dedicated

You can configure disk types, encryption keys, and OS images specific to Google Cloud Dedicated (GCD) deployments in the install-config.yaml file.

Example install-config.yaml file for GCD
apiVersion: v1
baseDomain: example.com
metadata:
  name: gcd-cluster
platform:
  gcp:
    projectID: eu0:my-project-id
    region: u-germany-northeast1
    defaultMachinePlatform:
      osDisk:
        diskType: hyperdisk-balanced
        encryptionKey:
          kmsKey:
            name: my-regional-key
            keyRing: my-key-ring
            location: u-germany-northeast1
            projectID: eu0:my-project-id
      osImage:
        project: eu0:my-project-id
        name: my-rhcos-image
publish: Internal
pullSecret: '{"auths": ...}'
sshKey: 'ssh-rsa AAAA...'

where:

baseDomain

Specifies the base domain for your cluster. Use a domain that you control.

platform.gcp.projectID

Specifies the Google Cloud project ID for your GCD environment. This value must include the domain-scoped prefix, such as eu0:my-project-id.

platform.gcp.defaultMachinePlatform.osDisk.diskType

Specifies the disk type for machine storage. For GCD, this value must be hyperdisk-balanced.

platform.gcp.defaultMachinePlatform.osDisk.encryptionKey.kmsKey.name

Specifies the name of your regional KMS key. In GCD, global keys are not supported.

platform.gcp.defaultMachinePlatform.osImage.project

Specifies the Google Cloud project ID where your custom FCOS image is stored. This value is required. The FCOS image is not pre-published in GCD regions.

publish

Specifies the cluster publishing strategy. For GCD, you must use Internal.

pullSecret

Specifies your Red Hat pull secret.

sshKey

Specifies your SSH public key for accessing cluster nodes.

Before you use a custom FCOS image, verify that it is compatible with your OKD version and has been properly imported into your GCD project.

Minimum resource requirements for cluster installation

Each created cluster must meet minimum requirements so that the cluster runs as expected.

Table 2. Minimum resource requirements
Machine Operating system vCPU Virtual RAM Storage Input/Output Per Second (IOPS)

Bootstrap

FCOS

4

16 GB

100 GB

300

Control plane

FCOS

4

16 GB

100 GB

300

Compute

FCOS

2

8 GB

100 GB

300

  • One vCPU is equivalent to one physical core when simultaneous multithreading (SMT), or Hyper-Threading, is not enabled. When enabled, use the following formula to calculate the corresponding ratio: (threads per core × cores) × sockets = vCPUs.

  • OKD and Kubernetes are sensitive to disk performance, and faster storage is recommended, particularly for etcd on the control plane nodes which require a 10 ms p99 fsync duration. Note that on many cloud platforms, storage size and IOPS scale together, so you might need to over-allocate storage volume to obtain sufficient performance.

  • As with all user-provisioned installations, if you choose to use Fedora compute machines in your cluster, you take responsibility for all operating system life cycle management and maintenance, including performing system updates, applying patches, and completing all other required tasks. Use of Fedora 7 compute machines is deprecated and has been removed in OKD 4.10 and later.

For OKD version 4.22, FCOS is based on Fedora version 9.8, which has the micro-architecture requirements. The following list contains the minimum instruction set architectures (ISA) that each architecture requires:

  • x86-64 architecture requires x86-64-v2 ISA

  • ARM64 architecture requires ARMv8.0-A ISA

  • ppc64le architecture requires IBM® Power9 ISA

  • s390x architecture requires IBM® z14 ISA

For more information, see Architectures in the Fedora documentation.

If an instance type for your platform meets the minimum requirements for cluster machines, it is supported to use in OKD.

Additional resources

Configuring the cluster-wide proxy during installation

Production environments can deny direct access to the internet and instead have an HTTP or HTTPS proxy available. You can configure a new OKD cluster to use a proxy by configuring the proxy settings in the install-config.yaml file.

Prerequisites

  • You have an existing install-config.yaml file.

  • You have reviewed the sites that your cluster requires access to and determined whether any of them need to bypass the proxy. By default, all cluster egress traffic is proxied, including calls to hosting cloud provider APIs. You added sites to the Proxy object’s spec.noProxy field to bypass the proxy if necessary.

    The Proxy object status.noProxy field is populated with the values of the networking.machineNetwork[].cidr, networking.clusterNetwork[].cidr, and networking.serviceNetwork[] fields from your installation configuration.

    For installations on Amazon Web Services (AWS), Google Cloud, Microsoft Azure, and OpenStack, the Proxy object status.noProxy field is also populated with the instance metadata endpoint (169.254.169.254).
Procedure

  1. Edit your install-config.yaml file and add the proxy settings. For example:

    apiVersion: v1
baseDomain: my.domain.com
proxy:
  httpProxy: http://<username>:<pswd>@<ip>:<port>
  httpsProxy: https://<username>:<pswd>@<ip>:<port>
  noProxy: example.com
additionalTrustBundle: |
    -----BEGIN CERTIFICATE-----
    <MY_TRUSTED_CA_CERT>
    -----END CERTIFICATE-----
additionalTrustBundlePolicy: <policy_to_add_additionalTrustBundle>
# ...

    where:

    proxy.httpProxy

    Specifies a proxy URL to use for creating HTTP connections outside the cluster. The URL scheme must be http.

    proxy.httpsProxy

    Specifies a proxy URL to use for creating HTTPS connections outside the cluster.

    proxy.noProxy

    Specifies a comma-separated list of destination domain names, IP addresses, or other network CIDRs to exclude from proxying. Preface a domain with . to match subdomains only. For example, .y.com matches x.y.com, but not y.com. Use * to bypass the proxy for all destinations.

    additionalTrustBundle

    If provided, the installation program generates a config map that is named user-ca-bundle in the openshift-config namespace to hold the additional CA certificates. If you provide additionalTrustBundle and at least one proxy setting, the Proxy object is configured to reference the user-ca-bundle config map in the trustedCA field. The Cluster Network Operator then creates a trusted-ca-bundle config map that merges the contents specified for the trustedCA parameter with the FCOS trust bundle. The additionalTrustBundle field is required unless the proxy’s identity certificate is signed by an authority from the FCOS trust bundle.

    additionalTrustBundlePolicy

    Specifies the policy that determines the configuration of the Proxy object to reference the user-ca-bundle config map in the trustedCA field. The allowed values are Proxyonly and Always. Use Proxyonly to reference the user-ca-bundle config map only when http/https proxy is configured. Use Always to always reference the user-ca-bundle config map. The default value is Proxyonly. Optional parameter.

    The installation program does not support the proxy readinessEndpoints field.

    If the installation program times out, restart and then complete the deployment by using the wait-for command of the installation program. For example:

    $ ./openshift-install wait-for install-complete --log-level debug

  2. Save the file and reference it when installing OKD.

    The installation program creates a cluster-wide proxy that is named cluster that uses the proxy settings in the provided install-config.yaml file. If no proxy settings are provided, a cluster Proxy object is still created, but it will have a nil spec.

    Only the Proxy object named cluster is supported, and no additional proxies can be created.

Alternatives to storing administrator-level secrets in the kube-system project

By default, administrator secrets are stored in the kube-system project. If you configured the credentialsMode parameter in the install-config.yaml file to Manual, you must manage long-term cloud credentials manually.

Manually creating long-term credentials

You can put the Cloud Credential Operator (CCO) into manual mode before OKD installation if the cloud identity and access management (IAM) APIs are not reachable, or if you prefer not to store an administrator-level credential secret in the cluster kube-system namespace.

Procedure

  1. Add the following granular permissions to the Google Cloud account that the installation program uses:

    • compute.machineTypes.list

    • compute.regions.list

    • compute.zones.list

    • dns.changes.create

    • dns.changes.get

    • dns.managedZones.create

    • dns.managedZones.delete

    • dns.managedZones.get

    • dns.managedZones.list

    • dns.networks.bindPrivateDNSZone

    • dns.resourceRecordSets.create

    • dns.resourceRecordSets.delete

    • dns.resourceRecordSets.list

  2. If you did not set the credentialsMode parameter in the install-config.yaml configuration file to Manual, modify the value as shown:

    Sample configuration file snippet
    apiVersion: v1
baseDomain: example.com
credentialsMode: Manual
# ...

  3. If you have not previously created installation manifest files, do so by running the following command:

    $ openshift-install create manifests --dir <installation_directory>

    where <installation_directory> is the directory in which the installation program creates files.

  4. Set a $RELEASE_IMAGE variable with the release image from your installation file by running the following command:

    $ RELEASE_IMAGE=$(./openshift-install version | awk '/release image/ {print $3}')

  5. Extract the list of CredentialsRequest custom resources (CRs) from the OKD release image by running the following command:

    $ oc adm release extract \
  --from=$RELEASE_IMAGE \
  --credentials-requests \
  --included \
  --install-config=<path_to_directory_with_installation_configuration>/install-config.yaml \
  --to=<path_to_directory_for_credentials_requests>

    where:

    --included

    Specifies only the manifests that your specific cluster configuration requires.

    <path_to_directory_with_installation_configuration>

    Specifies the location of the install-config.yaml file.

    <path_to_directory_for_credentials_requests>

    Specifies the path to the directory where you want to store the CredentialsRequest objects. If the specified directory does not exist, this command creates it.

    This command creates a YAML file for each CredentialsRequest object.

    Sample CredentialsRequest object
    apiVersion: cloudcredential.openshift.io/v1
kind: CredentialsRequest
metadata:
  name: <component_credentials_request>
  namespace: openshift-cloud-credential-operator
  ...
spec:
  providerSpec:
    apiVersion: cloudcredential.openshift.io/v1
    kind: GCPProviderSpec
    predefinedRoles:
    - roles/storage.admin
    - roles/iam.serviceAccountUser
    skipServiceCheck: true
  ...

  6. Create YAML files for secrets in the openshift-install manifests directory that you generated previously. The secrets must be stored using the namespace and secret name defined in the spec.secretRef for each CredentialsRequest object.

    Sample CredentialsRequest object with secrets
    apiVersion: cloudcredential.openshift.io/v1
kind: CredentialsRequest
metadata:
  name: <component_credentials_request>
  namespace: openshift-cloud-credential-operator
  ...
spec:
  providerSpec:
    apiVersion: cloudcredential.openshift.io/v1
      ...
  secretRef:
    name: <component_secret>
    namespace: <component_namespace>
  ...
    Sample Secret object
    apiVersion: v1
kind: Secret
metadata:
  name: <component_secret>
  namespace: <component_namespace>
data:
  service_account.json: <base64_encoded_gcp_service_account_file>

    Before upgrading a cluster that uses manually maintained credentials, you must ensure that the CCO is in an upgradeable state.

Installing the cluster

To deploy OKD in a sovereign cloud environment with strict data isolation, you can install a cluster on Google Cloud Dedicated by configuring and running the installation program.

Prerequisites

  • You prepared your GCD environment.

  • You obtained the OKD installation program.

  • You obtained the pull secret for your cluster.

  • You obtained your GCD region name and custom API endpoints from your local operating partner, such as Thales.

Procedure

  1. Create an install-config.yaml file. You can generate a template file by running the following command or manually create the file:

    $ ./openshift-install create install-config --dir <installation_directory>

    Replace <installation_directory> with the name of the directory that contains the installation files for your cluster.

    You cannot select a GCD sovereign cloud region by using the guided terminal prompts. You must manually edit the install-config.yaml file to configure GCD-specific parameters.

  2. Edit the install-config.yaml file to configure it for your GCD environment:

    The installation program automatically detects the GCD sovereign cloud environment based on the region and the domain-scoped project ID in your credential file. You do not need to manually configure API endpoints.

    1. Specify the GCD region:

      platform:
  gcp:
    region: <gcd_region_name>

    2. Configure the cluster for private networking:

      publish: Internal

      where:

      publish

      Specifies the cluster publishing strategy. For GCD, you must use Internal. Public clusters are not supported.

    3. Optional: If you need to provide a custom Fedora CoreOS (FCOS) image, add the compute platform image field:

      platform:
  gcp:
    defaultMachinePlatform:
      osImage:
        project: <project_id>
        name: <image_name>

      where:

      platform.gcp.defaultMachinePlatform.osImage.project

      Specifies the Google Cloud project ID where your custom FCOS image is stored.

      platform.gcp.defaultMachinePlatform.osImage.name

      Specifies the name of your custom FCOS image.

      You must specify a custom FCOS image. The FCOS image is not pre-published in GCD regions. If you do not provide the osImage.name and osImage.project fields, the installation fails.

    4. Optional: Configure Customer-Managed Encryption Keys (CMEK) if required for regulatory compliance:

      When you use CMEK with GCD, you must use regional keys. Global keys are not supported.

      		 
      platform:
  gcp:
    defaultMachinePlatform:
      osDisk:
        encryptionKey:
          kmsKey:
            name: <regional_key_name>
            keyRing: <key_ring_name>
            location: <region_name>
            projectID: <project_id>

      where:

      platform.gcp.defaultMachinePlatform.osDisk.encryptionKey.kmsKey.name

      Specifies a regional KMS key, not a global key.

    5. Configure storage to use hyperdisk-balanced storage:

      platform:
  gcp:
    defaultMachinePlatform:
      osDisk:
        diskType: hyperdisk-balanced

      where:

      platform.gcp.defaultMachinePlatform.osDisk.diskType

      Specifies the disk type for machine storage. GCD supports only hyperdisk-balanced storage. Other disk types are not available.

  3. Back up the install-config.yaml file by running the following command:

    $ cp <installation_directory>/install-config.yaml <installation_directory>/install-config.yaml.backup

    The install-config.yaml file is consumed during the installation process. If you want to reuse the file, you must back it up.

  4. Generate the Kubernetes manifests for the cluster by running the following command:

    $ ./openshift-install create manifests --dir <installation_directory>

  5. Deploy the OKD cluster by running the following command:

    $ ./openshift-install create cluster --dir <installation_directory> --log-level=info

    If you provided correct configuration values, the cluster creation process should complete successfully. However, the installation program might fail if it cannot validate your custom API endpoints or if there are connectivity issues with your GCD environment.

    When the cluster deployment is complete, the installation program displays information about accessing your cluster, including a link to the web console and credentials for the kubeadmin user.

    To verify that the cluster Operators are running, set the KUBECONFIG environment variable and check the cluster Operator status by running the following commands:

    $ export KUBECONFIG=<installation_directory>/auth/kubeconfig
    $ oc get clusteroperators

    Verify that all cluster Operators show AVAILABLE=True, PROGRESSING=False, and DEGRADED=False.

Logging in to the cluster by using the CLI

To log in to your cluster as the default system user, export the kubeconfig file. This configuration enables the CLI to authenticate and connect to the specific API server created during OKD installation.

The kubeconfig file is specific to a cluster and is created during OKD installation.

Prerequisites

  • You deployed an OKD cluster.

  • You installed the OpenShift CLI (oc).

Procedure

  1. Export the kubeadmin credentials by running the following command:

    $ export KUBECONFIG=<installation_directory>/auth/kubeconfig

    where:

    <installation_directory>

    Specifies the path to the directory that stores the installation files.

  2. Verify you can run oc commands successfully using the exported configuration by running the following command:

    $ oc whoami
    Example output
    system:admin
Next steps

  • "Customize your cluster"

  • "Remote health reporting"

Additional resources
Additional resources

Additional resources