Preparing an Agent-based installed cluster for the multicluster engine for Kubernetes - Installing an on-premise cluster with the Agent-based Installer | Installing

Prerequisites
Preparing an Agent-based cluster deployment for the multicluster engine for Kubernetes Operator while disconnected
Preparing an Agent-based cluster deployment for the multicluster engine for Kubernetes Operator while connected

You can install the multicluster engine Operator and deploy a hub cluster with the Agent-based OpenShift Container Platform Installer. The following procedure is partially automated and requires manual steps after the initial cluster is deployed.

Prerequisites

You have read the following documentation:
You have access to the internet to obtain the necessary container images.
You have installed the OpenShift CLI (oc).
If you are installing in a disconnected environment, you must have a configured local mirror registry for disconnected installation mirroring.

Preparing an Agent-based cluster deployment for the multicluster engine for Kubernetes Operator while disconnected

You can mirror the required OpenShift Container Platform container images, the multicluster engine Operator, and the Local Storage Operator (LSO) into your local mirror registry in a disconnected environment. Ensure that you note the local DNS hostname and port of your mirror registry.

To mirror your OpenShift Container Platform image repository to your mirror registry, you can use either the oc adm release image or oc mirror command. In this procedure, the oc mirror command is used as an example.

Procedure

Create an <assets_directory> folder to contain valid install-config.yaml and agent-config.yaml files. This directory is used to store all the assets.

To mirror an OpenShift Container Platform image repository, the multicluster engine, and the LSO, create a ImageSetConfiguration.yaml file with the following settings:

Example ImageSetConfiguration.yaml

  kind: ImageSetConfiguration
  apiVersion: mirror.openshift.io/v1alpha2
  archiveSize: 4 (1)
  storageConfig: (2)
    imageURL: <your-local-registry-dns-name>:<your-local-registry-port>/mirror/oc-mirror-metadata (3)
    skipTLS: true
  mirror:
    platform:
      architectures:
        - "amd64"
      channels:
        - name: stable-4.14 (4)
          type: ocp
    additionalImages:
      - name: registry.redhat.io/ubi9/ubi:latest
    operators:
      - catalog: registry.redhat.io/redhat/redhat-operator-index:v4.14 (5)
        packages: (6)
          - name: multicluster-engine (7)
          - name: local-storage-operator (8)

1	Specify the maximum size, in GiB, of each file within the image set.
2	Set the back-end location to receive the image set metadata. This location can be a registry or local directory. It is required to specify `storageConfig` values.
3	Set the registry URL for the storage backend.
4	Set the channel that contains the OpenShift Container Platform images for the version you are installing.
5	Set the Operator catalog that contains the OpenShift Container Platform images that you are installing.
6	Specify only certain Operator packages and channels to include in the image set. Remove this field to retrieve all packages in the catalog.
7	The multicluster engine packages and channels.
8	The LSO packages and channels.

This file is required by the oc mirror command when mirroring content.

To mirror a specific OpenShift Container Platform image repository, the multicluster engine, and the LSO, run the following command:
```
$ oc mirror --dest-skip-tls --config ocp-mce-imageset.yaml docker://<your-local-registry-dns-name>:<your-local-registry-port>
```

Update the registry and certificate in the install-config.yaml file:

Example imageContentSources.yaml

  imageContentSources:
    - source: "quay.io/openshift-release-dev/ocp-release"
      mirrors:
        - "<your-local-registry-dns-name>:<your-local-registry-port>/openshift/release-images"
    - source: "quay.io/openshift-release-dev/ocp-v4.0-art-dev"
      mirrors:
        - "<your-local-registry-dns-name>:<your-local-registry-port>/openshift/release"
    - source: "registry.redhat.io/ubi9"
      mirrors:
        - "<your-local-registry-dns-name>:<your-local-registry-port>/ubi9"
    - source: "registry.redhat.io/multicluster-engine"
      mirrors:
        - "<your-local-registry-dns-name>:<your-local-registry-port>/multicluster-engine"
    - source: "registry.redhat.io/rhel8"
      mirrors:
        - "<your-local-registry-dns-name>:<your-local-registry-port>/rhel8"
    - source: "registry.redhat.io/redhat"
      mirrors:
        - "<your-local-registry-dns-name>:<your-local-registry-port>/redhat"

Additionally, ensure your certificate is present in the additionalTrustBundle field of the install-config.yaml.

Example install-config.yaml

additionalTrustBundle: |
  -----BEGIN CERTIFICATE-----
  zzzzzzzzzzz
  -----END CERTIFICATE-------

The oc mirror command creates a folder called oc-mirror-workspace with several outputs. This includes the imageContentSourcePolicy.yaml file that identifies all the mirrors you need for OpenShift Container Platform and your selected Operators.

Generate the cluster manifests by running the following command:
```
$ openshift-install agent create cluster-manifests
```
This command updates the cluster manifests folder to include a mirror folder that contains your mirror configuration.

Preparing an Agent-based cluster deployment for the multicluster engine for Kubernetes Operator while connected

Create the required manifests for the multicluster engine Operator, the Local Storage Operator (LSO), and to deploy an agent-based OpenShift Container Platform cluster as a hub cluster.

Procedure

Create a sub-folder named openshift in the <assets_directory> folder. This sub-folder is used to store the extra manifests that will be applied during the installation to further customize the deployed cluster. The <assets_directory> folder contains all the assets including the install-config.yaml and agent-config.yaml files.

The installer does not validate extra manifests.

For the multicluster engine, create the following manifests and save them in the <assets_directory>/openshift folder:

Example mce_namespace.yaml

  apiVersion: v1
  kind: Namespace
  metadata:
    labels:
      openshift.io/cluster-monitoring: "true"
    name: multicluster-engine

Example mce_operatorgroup.yaml

  apiVersion: operators.coreos.com/v1
  kind: OperatorGroup
  metadata:
    name: multicluster-engine-operatorgroup
    namespace: multicluster-engine
  spec:
    targetNamespaces:
    - multicluster-engine

Example mce_subscription.yaml

  apiVersion: operators.coreos.com/v1alpha1
  kind: Subscription
  metadata:
    name: multicluster-engine
    namespace: multicluster-engine
  spec:
    channel: "stable-2.3"
    name: multicluster-engine
    source: redhat-operators
    sourceNamespace: openshift-marketplace

You can install a distributed unit (DU) at scale with the Red Hat Advanced Cluster Management (RHACM) using the assisted installer (AI). These distributed units must be enabled in the hub cluster. The AI service requires persistent volumes (PVs), which are manually created.

For the AI service, create the following manifests and save them in the <assets_directory>/openshift folder:

Example lso_namespace.yaml

  apiVersion: v1
  kind: Namespace
  metadata:
    annotations:
      openshift.io/cluster-monitoring: "true"
    name: openshift-local-storage

Example lso_operatorgroup.yaml

  apiVersion: operators.coreos.com/v1
  kind: OperatorGroup
  metadata:
    name: local-operator-group
    namespace: openshift-local-storage
  spec:
    targetNamespaces:
      - openshift-local-storage

Example lso_subscription.yaml

  apiVersion: operators.coreos.com/v1alpha1
  kind: Subscription
  metadata:
    name: local-storage-operator
    namespace: openshift-local-storage
  spec:
    installPlanApproval: Automatic
    name: local-storage-operator
    source: redhat-operators
    sourceNamespace: openshift-marketplace

After creating all the manifests, your filesystem must display as follows:

Example Filesystem

<assets_directory>
    ├─ install-config.yaml
    ├─ agent-config.yaml
    └─ /openshift
        ├─ mce_namespace.yaml
        ├─ mce_operatorgroup.yaml
        ├─ mce_subscription.yaml
        ├─ lso_namespace.yaml
        ├─ lso_operatorgroup.yaml
        └─ lso_subscription.yaml

Create the agent ISO image by running the following command:

$ openshift-install agent create image --dir <assets_directory>

When the image is ready, boot the target machine and wait for the installation to complete.

To monitor the installation, run the following command:

$ openshift-install agent wait-for install-complete --dir <assets_directory>

To configure a fully functional hub cluster, you must create the following manifests and manually apply them by running the command $ oc apply -f <manifest-name>. The order of the manifest creation is important and where required, the waiting condition is displayed.

For the PVs that are required by the AI service, create the following manifests:

  apiVersion: local.storage.openshift.io/v1
  kind: LocalVolume
  metadata:
   name: assisted-service
   namespace: openshift-local-storage
  spec:
   logLevel: Normal
   managementState: Managed
   storageClassDevices:
     - devicePaths:
         - /dev/vda
         - /dev/vdb
       storageClassName: assisted-service
       volumeMode: Filesystem

Use the following command to wait for the availability of the PVs, before applying the subsequent manifests:

$ oc wait localvolume -n openshift-local-storage assisted-service --for condition=Available --timeout 10m

The `devicePath` is an example and may vary depending on the actual hardware configuration used.

Create a manifest for a multicluster engine instance.

Example MultiClusterEngine.yaml

  apiVersion: multicluster.openshift.io/v1
  kind: MultiClusterEngine
  metadata:
    name: multiclusterengine
  spec: {}

Create a manifest to enable the AI service.

Example agentserviceconfig.yaml

  apiVersion: agent-install.openshift.io/v1beta1
  kind: AgentServiceConfig
  metadata:
    name: agent
    namespace: assisted-installer
  spec:
   databaseStorage:
    storageClassName: assisted-service
    accessModes:
    - ReadWriteOnce
    resources:
     requests:
      storage: 10Gi
   filesystemStorage:
    storageClassName: assisted-service
    accessModes:
    - ReadWriteOnce
    resources:
     requests:
      storage: 10Gi

Create a manifest to deploy subsequently spoke clusters.

Example clusterimageset.yaml

  apiVersion: hive.openshift.io/v1
  kind: ClusterImageSet
  metadata:
    name: "4.14"
  spec:
    releaseImage: quay.io/openshift-release-dev/ocp-release:4.14.0-x86_64

Create a manifest to import the agent installed cluster (that hosts the multicluster engine and the Assisted Service) as the hub cluster.

Example autoimport.yaml

  apiVersion: cluster.open-cluster-management.io/v1
  kind: ManagedCluster
  metadata:
   labels:
     local-cluster: "true"
     cloud: auto-detect
     vendor: auto-detect
   name: local-cluster
  spec:
   hubAcceptsClient: true

Wait for the managed cluster to be created.

$ oc wait -n multicluster-engine managedclusters local-cluster --for condition=ManagedClusterJoined=True --timeout 10m

Verification

To confirm that the managed cluster installation is successful, run the following command:

$ oc get managedcluster
NAME            HUB ACCEPTED   MANAGED CLUSTER URLS             JOINED   AVAILABLE  AGE
local-cluster   true           https://<your cluster url>:6443   True     True       77m

Additional resources

The Local Storage Operator

Preparing an Agent-based installed cluster for the multicluster engine for Kubernetes Operator

Prerequisites

Preparing an Agent-based cluster deployment for the multicluster engine for Kubernetes Operator while disconnected

Preparing an Agent-based cluster deployment for the multicluster engine for Kubernetes Operator while connected