Installing stand-alone AI Manager (OLM UI)

Complete the following steps to install IBM Watson AIOps AI Manager.

Prerequisites

Before installing the Watson AIOps AI Manager operator, you must complete the tasks listed in Preparing to install the Watson AIOps or AI Manager operator.

Procedures

Add the AI Manager catalog source
Install the AI Manager operator
Create an AI Manager instance

1. Add the catalog source

Log in to your Red Hat OpenShift Container Platform cluster console.
Click Administration > Cluster Settings, then select the OperatorHub configuration resource under the Global Configurations tab.
Click Sources > Create Catalog Source. Provide the IBM Watson AI Manager catalog source name ibm-watson-aimanager-catalog, the image URL docker.io/ibmcom/ibm-watson-ai-manager-operator-catalog:latest, and select Cluster-wide catalog source, then click Create.
The IBM Watson AIOps AI Manager catalog source appears. Refresh the screen after a few minutes, and ensure that the # of operators count is 1.
Click the IBM Watson AIOps AI Manager source, then click YAML. If the spec object in your catalog source YAML contains updateStrategy, continue with installing Event Manager. If the spec object does not include updateStrategy, add the following code underneath spec to your catalog source YAML, then click Save:
```
spec:
  updateStrategy:
      registryPoll:
          interval: 45m
```

2. Install the operator

Prerequisite: Make sure that you created both of the image pull secrets as defined in Create the entitled registry secret.

Click Operators > OperatorHub, and search for the IBM Watson AIOps AI Manager operator. Click Install.
Select the namespace that you created when you prepared for your AI Manager installation to install the operator. Do not use namespaces that are owned by Kubernetes or OpenShift, such as kube-system or default.
Select Automatic as the Approval Strategy.
Click Install.
Click Operators > Installed Operators, and view the IBM Watson AIOps AI Manager operator. It takes a few minutes to install. Ensure that the status of the installed Watson AIOps operator is Succeeded.
Ensure that the IBM Watson AIOps AI Manager operator is set to the latest update channel:
1. Click the IBM Watson AIOps AI Manager operator, then click Subscription.
2. Click the name in the Channel field in the Subscription Details table, then select the latest stable channel.
3. Click Save.

3. Create an AI Manager instance

From the OpenShift console, click Operators > Installed Operators, and select the Watson AIOps AI Manager operator. In the Provided APIs section of the Overview tab, click Create Instance in the AI Manager operator.
Configure the properties for the AI Manager deployment in either the Form View or YAML View of the OpenShift console:
1. Enter the following values if you are configuring your instance in the Form View:
  - Existing Pull Secret: Enter your cp.stg.icr.io secret. Note that if you are not creating your instance for a private registry, leave the Repo value blank.
  - Storage > Storage Class Name: The name of the storage class that you installed on your cluster. For example, portworx-sc.
  - License > License accept: Toggle the license to True to accept the AI Manager license.
2. Enter the following values if you are configuring your instance in the YAML View:
  - existingPullSecret > name: Enter your cp.stg.icr.io secret. Note that if you are not creating your instance for a private registry, leave the repovalue blank.
  - storage > storageClassName: The name of the storage class that you installed on your cluster. For example, portworx-sc.
  - license > accept: Change the value to true to accept the AI Manager license.
Notes:
- If you want to change the Storage Size by clicking the properties twistie for Flink, Kafka, MinIO, or PostrgeSQL, click Storage > Storage Size. You must enter only integer values, for example 20Gi, 100Gi, and so on. Entering other values blocks the creation of your AI Manager instance.
- If you want to deploy the AI Manager to Red Hat OpenShift Kubernetes Service, you must add the following configurations after zen, but at the same level in the YAML view of the AI Manager deployment:
```
aimanagerValues:
  postgres:
    keeper:
      resources:
        limits:
          cpu: 500m
          memory: 512Mi
        requests:
          cpu: 100m
          memory: 256Mi
    livenessProbe:
      failureThreshold: 5
      initialDelaySeconds: 300
      periodSeconds: 15
      successThreshold: 1
      timeoutSeconds: 15
    readinessProbe:
      failureThreshold: 5
      initialDelaySeconds: 300
      periodSeconds: 15
      successThreshold: 1
      timeoutSeconds: 15
  sentinel:
    resources:
      limits:
        cpu: 500m
        memory: 512Mi
      requests:
        cpu: 100m
        memory: 256Mi
```
Click Create.
Under the All Instances tab, your AI Manager instance appears. View the status of each of the updates on the installation. When the instances status shows OK, the deployment is ready.
If you want to enable or disable a feature after your installation, you can edit the AI Manager instance by running the following command:
```
oc edit aimanager.aiops.watson.ibm.com <ai-manager-instance-name>
```
Where <ai-manager-instance-name> is the name of the instance that you want to edit. You can then select to enable or disable the feature.

Note: When you disable the features after installation, the resource is not automatically deleted. To check whether a feature is deleted or not, you need to check the operator log.