Installation via OLM

The Operator Lifecycle Manager (OLM) provides a framework for installing, updating, and managing the lifecycle of operators and their services.

Prerequisites
Installing the CatalogSource
Installing the DataPower Operator
References

Prerequisites

The DataPower Operator currently supports installation via OLM in OCP clusters running version 4.6 or higher.

Coexistence

When installing an operator via OLM, you have the option to choose its Installation Mode. There are two options:

All namespaces on the cluster (AllNamespaces)
A specific namespace on the cluster (OwnNamespace)

If installed in AllNamespaces mode, the Operator will use a ClusterRole and ClusterRoleBinding and thus have cluster-wide scope to manage DataPower resources across all namespaces. If installed in OwnNamespace mode, the operator will use a Role and RoleBinding as its primary access (isolated to the namespace it’s installed in), with a limited set of ClusterRole permissions (see Cluster-scope permissions).

Warning: You should not install the Operator in more than one mode. If AllNamespaces is chosen, do not subsequently install a second instance in OwnNamespace mode.

Warning: If installing more than one instance of the Operator in OwnNamespace mode across multiple namespaces, they MUST be installed using the same channel / version.

Installing the CatalogSource

To enable the installation of the DataPower Operator, a CatalogSource must be created in the openshift-marketplace.

The below YAML snippet can be used to create the CatalogSource resource, which references the datapower-operator-catalog image from IBM Container Registry. The use of the latest tag here will allow OLM to poll IBM Container Registry and pull new catalog images when they become available, enabling over-the-air updates.

apiVersion: operators.coreos.com/v1alpha1
kind: CatalogSource
metadata:
  name: ibm-datapower-operator-catalog
  namespace: openshift-marketplace
spec:
  displayName: IBM DataPower Operator
  publisher: IBM
  sourceType: grpc

You can create this resource manually either via the OpenShift Console UI, or by using the oc CLI.

Using the OpenShift Console

To creating this resource using the OpenShift Console, use the following steps.

Navigate to your OpenShift Console UI.
In the top-right of the UI, on the header bar, click the Import button (+) to import YAML.
Copy and paste the above YAML snippet into the editor.
Click the Create button to create the resource.

Using the `oc` CLI

To create this resource using the oc CLI, use the following steps.

Create a YAML file containing the above YAML snippet.

Use the oc apply command to apply the YAML resource.

oc apply -f ibm-datapower-operator-catalog.yaml

Validating the CatalogSource is installed

To validate that the CatalogSource resource was installed correctly, use the following steps.

Navigate to the OpenShift Console UI.
On the left panel, expand the Operators section.
Select OperatorHub.
At the top of the OperatorHub section, enter DataPower into the Filter search box.
A tile should be shown titled IBM DataPower Gateway.

Installing the DataPower Operator

To install an instance of the DataPower Operator, use the following steps.

Use the steps above to locate the IBM DataPower Gateway tile in the OperatorHub UI.
Select the IBM DataPower Gateway tile. A panel to the right should appear.
Click the Install button on the right panel.
Under Installation Mode select your desired installation mode.
Select the desired Update Channel.
Select the desired Approval Strategy.
The Approval Strategy is what determines if the DataPower Operator will automatically update when new releases become available within the selected channel. If Automatic is selected, over-the-air updates will occur automatically as they become available. If Manual is selected, an administrator would need to approve each update as it becomes available through OLM.
Click the Subscribe button to install the DataPower Operator.

References

The below references might be helpful if you wish to learn more about OLM.

Edit this page on GitHub

Install: Prerequisites

Install: Installation via CASE