Installing IBM Watson ACD

Overview

1. Install the IBM operator catalog: This will deploy the catalog from which IBM operators, including ACD, can be installed.
2. Install the ACD operator: This will deploy the operator that can be used to install and manage your ACD instances.
3. Install one or more replicas of ACD by using the ACD operator.

Before you begin

• Plan for your installation, such as preparing for persistent storage, considering security options, and planning for performance and capacity.
• Obtain an entitlement key and verify registry access to the IBM Entitled Registry. Note that customers must use their IBMid to log in to their myibm account. The customer must request the entitlement key so the ownership and management of the entitlement stays with them.
• Set up your environment according to the prerequisites, including setting up your OpenShift Container Platform.
• Obtain the connection details for your OpenShift Container Platform cluster from your administrator.
• Set up your project and project dependencies if required for your environment. If using the CLI to install the ACD operator and ACD service, you will need the cloudctl command line utility.

Verifying IBM Entitled Registry access

docker login cp.icr.io --username cp --password <your entitlement key>
Copy to clipboardCopied!

Air-gapped (disconnected) installation

Non air-gapped (connected) installation

IBM Entitled Registry pull secret

1. Added to the OpenShift global pull secrets
2. Added to the ACD operand service account

Option 1: OpenShift global pull secret installation

1. Use the RedHat OpenShift Container Platform web console and select Workloads -> Secrets in the openshift-config project.

2. Select the pull-secret object from the list of secrets.

3. Select Actions -> Edit Secret to bring up the secret editor for this secret.

4. Edit an existing credential for your registry if it already exists or Select Add credentials at the bottom and fill in the new pull secret credentials.

   

1. Extract the current global image pull secret from the cluster into a file in the current directory named .dockerconfigjson:

   oc extract secret/pull-secret --namespace openshift-config --to=.

2. Create a base64 encoded string with the registry userid and password as it aligns with your access method.

   printf "cp:<acd registry key>" | base64

3. Edit the .dockerconfigjson file and ADD a new JSON object to the existing auths object with the credentials for the IBM Entitled Registry. For example:

   "cp.icr.io": {
       "auth": "aWFtYXBpaxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxcGFzc3dvcmQ=",
       "email": "xxx@nomail.relay.ibm.com"
   }
   Copy to clipboardCopied!

4. Update the global image pull secret with the updated credentials:

   oc set data secret/pull-secret --namespace openshift-config --from-file=.dockerconfigjson

5. Monitor the node status using the command:

   oc get nodes

6. When the nodes are finished restarting, your cluster is now ready to pull images from the registry.

Option 2: Service account pull secret installation

1. Create a secret

   kubectl create secret docker-registry ibm-entitlement-key \
       --docker-server=cp.icr.io \
       --docker-username=<username> \
       --docker-password=<password> \
       --docker-email=<email_address> \
       --namespace=<namespace>
   Copy to clipboardCopied!

   • <username> is the username for the entitled registry. This should be cp when using a myibm entitlement key.
   • <password> is the password for the entitled registry. This should be the entitlement key from myibm.

2. After the ACD operand has been installed, the service account must be patched to point to the secret.

   NOTE: If using the current release of the ACD Container Edition, this ibm-entitlement-key pull secret is already defined in the operand service account so the patch step is no longer necessary.

   kubectl patch serviceaccount ibm-wh-acd-operand \
       --namespace <namespace> \
       --patch '{"imagePullSecrets": [{"name": "ibm-entitlement-key"}]}'
   Copy to clipboardCopied!

3. Then the ACD operand pods must be restarted

   kubectl delete pods \
       --namespace <namespace> \
       --all
   Copy to clipboardCopied!

Installing the IBM operator catalog

1. Create a file for the IBM operator catalog source with the following content, and save as IBMCatalogSource.yaml:

   apiVersion: operators.coreos.com/v1alpha1
   kind: CatalogSource
   metadata:
      name: ibm-operator-catalog
      namespace: openshift-marketplace
   spec:
      displayName: "IBM Operator Catalog"
      publisher: IBM
      sourceType: grpc
   Copy to clipboardCopied!

2. Log in to your Red Hat OpenShift Container Platform as a cluster administrator by using the oc CLI.

3. Apply the source by using the following command:

   oc apply -f IBMCatalogSource.yaml

Installing the ACD operator

Install the ACD operator using the web console

1. Log in to the OpenShift Container Platform web console using your login credentials.
2. Expand the Operators dropdown and select OperatorHub to open the OperatorHub dashboard.
3. Select the project you want to use as the target namespace for your ACD deployment.
4. In the All Items search box enter ACD to locate the operator title.
5. Click the ACD tile to open the install side panel.
6. Click the Install button to open the Create Operator Subscription dashboard.
7. Select the chosen installation mode that suits your requirements. If the installation mode is A specific namespace on the cluster, select the target namespace you created previously.
8. Select the approval strategy that suits your requirements. If set to Automatic, the Subscription resource uses Operator Lifecycle Manager (OLM) to manage and upgrade the operator to ensure that the latest version is always running in the cluster. With Manual approval set, a project administrator must manually approve the install plan to enable the upgrade. See the upgrading section for more details.
9. Click Install to begin the installation.

Install the ACD operator using cloudctl

cloudctl case launch \
    --case case/ibm-wh-acd \
    --namespace <namespace> \
    --inventory clinicalDataAnnotatorOperatorSetup \
    --action installOperator \
    --tolerance 1
Copy to clipboardCopied!

Installing the ACD service

Install the ACD service by using the web console

1. Log in to the OpenShift Container Platform web console using your login credentials.
2. Expand the Operators dropdown and select Installed Operators to open the Installed Operators page.
3. Expand the Project dropdown and select the project the operator is installed in. Select the Annotator for Clinical Data operator link in the Name column. If the operator is not shown, it is either not installed or not available for the selected namespace.
4. In the Operator Details dashboard, click the Annotator for Clinical Data tab.
5. Click the Create Acd button to open the Create Acd panel. You can use this panel to define an Acd custom resource.
   • From here, you can install by using the Form view. For more advanced configurations or to install an instance using default configuration, see installing by using the YAML view.

Install the ACD service using cloudctl

cloudctl case launch \
    --case case/ibm-wh-acd \
    --namespace <namespace> \
    --inventory clinicalDataAnnotatorOperator \
    --action applyCustomResources \
    --tolerance 1
Copy to clipboardCopied!

cloudctl case launch \
    --case case/ibm-wh-acd \
    --namespace <namespace> \
    --inventory clinicalDataAnnotatorOperator \
    --action applyCustomResources \
    --tolerance 1 \
    --args "--backend cos --bucket <bucket> --endpointUrl <endpoint> --location <location>"
Copy to clipboardCopied!