Installing Watson Machine Learning on-premises

Watson Machine Learning provides a full range of tools and services so that you can build, train, and deploy Machine Learning models. Choose the tool with the level of automation or autonomy that matches your needs, from a fully automated process to writing your own code.

For more information about Watson Machine Learning, see Watson Machine Learning on Cloud Pak for Data .

Tip: This task maps to the following Ansible role, when you specify wml as the service name: cp4d_service. For more information, see IBM Maximo Application Suite installation with Ansible collection.

Installing by using the Red Hat® OpenShift® Container Platform web console

Procedure

In the banner, click Import YAML (

). Enter the following YAML.


---
apiVersion: operators.coreos.com/v1alpha1
kind: Subscription
metadata:
  labels:
    app.kubernetes.io/instance: ibm-cpd-wml-operator-subscription
    app.kubernetes.io/managed-by: ibm-cpd-wml-operator
    app.kubernetes.io/name: ibm-cpd-wml-operator-subscription
  name: ibm-cpd-wml-operator-subscription
  namespace: ibm-common-services
spec:
    channel: v1.1
    installPlanApproval: Automatic
    name: ibm-cpd-wml-operator
    source: ibm-operator-catalog
    sourceNamespace: openshift-marketplace

Click the Create to create the Watson Machine Learning subscription.
On the Workloads > Deployments page, switch to the ibm-common-services project and wait for the Watson Machine Learning pod resource ibm-cpd-wml-operator to indicate it is in Running state.

In the banner, click Import YAML (

). Enter the following YAML.


---
apiVersion: wml.cpd.ibm.com/v1beta1
kind: WmlBase
metadata:
  name: wml-cr
  namespace: cpd-services
  labels:
    app.kubernetes.io/instance: wml-cr
    app.kubernetes.io/managed-by: ibm-cpd-wml-operator
    app.kubernetes.io/name: ibm-cpd-wml-operator
spec:
  ignoreForMaintenance: false
  scaleConfig: small
  license:
    accept: true
    license: Standard
  version: 4.0.8
  storageVendor: ocs
  storageClass: ocs-storagecluster-cephfs

Click the Create button to create the WML instance.
On the Home > Search page, select the cpd-services project and search for resource type wmlbase. In the search results, click the wml-cr. Show the YAML for the wml-cr instance and wait until you see wmlStatus: Completed in the status section. This can take 45 minutes.

Installing by using the OpenShift command-line interface (CLI)

Procedure

Install the Watson Machine Learning subscription YAML file wml-subscription.yaml:


---
apiVersion: operators.coreos.com/v1alpha1
kind: Subscription
metadata:
  labels:
    app.kubernetes.io/instance: ibm-cpd-wml-operator-subscription
    app.kubernetes.io/managed-by: ibm-cpd-wml-operator
    app.kubernetes.io/name: ibm-cpd-wml-operator-subscription
  name: ibm-cpd-wml-operator-subscription
  namespace: ibm-common-services
spec:
    channel: v1.1
    installPlanApproval: Automatic
    name: ibm-cpd-wml-operator
    source: ibm-operator-catalog
    sourceNamespace: openshift-marketplace

Apply the wml-subscription.yaml file to the OCP cluster:
```
oc apply -f wml-subscription.yaml
```

Verify the WML Subscription, csv and deployment pod:


oc get sub -n ibm-common-services ibm-cpd-wml-operator-subscription \
> -o jsonpath='{.status.installedCSV} {"\n"}'

Sample output

ibm-cpd-wml-operator.v1.1.7


oc get csv -n ibm-common-services ibm-cpd-wml-operator.v1.1.7 \
> -o jsonpath='{ .status.phase } : { .status.message} {"\n"}'

Sample output

Succeeded : install strategy completed with no errors


oc get deployments -n ibm-common-services -l olm.owner="ibm-cpd-wml-operator.v1.1.7" \
> -o jsonpath="{.items[0].status.availableReplicas} {'\n'}"

Sample output

Create a custom resource YAML file wml-cr.yaml to install wmlbase instance:


---
apiVersion: wml.cpd.ibm.com/v1beta1
kind: WmlBase
metadata:
  name: wml-cr
  namespace: cpd-services
  labels:
    app.kubernetes.io/instance: wml-cr
    app.kubernetes.io/managed-by: ibm-cpd-wml-operator
    app.kubernetes.io/name: ibm-cpd-wml-operator
spec:
  ignoreForMaintenance: false
  scaleConfig: small
  license:
    accept: true
    license: Standard
  version: 4.0.8
  storageVendor: ocs
  storageClass: ocs-storagecluster-cephfs

Apply the wml-cr.yaml file to the OCP cluster:
```
oc apply -f wml-cr.yaml
```
Verify that the Watson Machine Learning is installed and Watson Machine Learning instance status is Completed:
```
oc get WmlBase wml-cr -o jsonpath='{.status.wmlStatus} {"\n"}' -n cpd-services 
```