Cloud Pak for Data Deployment¶

CPD Deployment on IBM Cloud¶

To deploy IBM Cloud Pak for Data on an OpenShift cluster, we will use the IBM Cloud Native Toolkit GitOps Framework. There are only five steps to do:

Prereqs - Make sure you have a RedHat OpenShift cluster and you are able to use the RedHat OpenShift CLI against it.
Sealed Secrets - Provide the private key used to seal the secrets provided with the API Connect GitOps repository.
RedHat OpenShift GitOps Operator - Install the RedHat OpenShift GitOps operator which provides the GitOps tools needed for installing and managing IBM Cloud Pak for Data instances using the GitOps approach already explained.
IBM Cloud Pak for Data - Deploy an instance of IBM Cloud Pak for Data on the RedHat OpenShift cluster.
IBM Cloud Pak for Data UI - Validate the installation of your IBM Cloud Pak for Data instance by making sure you are able to log into the IBM Cloud Pak for Data user interface.

1 - Prereqs¶

Obtain a clean RedHat OpenShift cluster where nothing else has been installed on top. This RedHat OpenShift cluster must be composed of five worker nodes where three of these will be entirely dedicated to OpenShift Data Foundation (ODF). The storage nodes must be 16 CPUs and 64 GB RAM at least.
Once your RedHat OpenShift cluster is available on your IBM Cloud dashboard, you must install the OpenShift Data Foundation add-on:
1. From the OpenShift clusters console, select the cluster where you want to install the add-on.
2. On the cluster Overview page, click Add-ons.
3. On the OpenShift Data Foundation card, click Install.
On the cluster Nodes page, select one node per availability zone and write down it's IP address.
Log into your RedHat OpenShift web console using the button on the top right corner of your cluster dashboard on IBM Cloud.
Go to the Compute --> Nodes section on the navigation index on the left hand side and for each of the nodes you picked in the previous step:
1. Click on the node and in its Node Details dashboard, click on the Actions drop down menu that appears on the top right corner and select Edit Labels.
2. Type node-role.kubernetes.io/storage=true and hit enter to add that label to the node. Click Save.
Once you have completed the process above for the three nodes you selected to be entirely dedicated to storage, you should see such role in the Compute --> Nodes section of your RedHat OpenShift cluster.
Log into your RedHat OpenShift cluster through the RedHat OpenShift CLI in order to execute commands to interact with it through your terminal.

2 - Sealed Secrets¶

Create the sealed-secrets project. This project will host the Sealed Secrets operator that will allow us to decrypt sealed secrets stored in GitHub.
```
oc new-project sealed-secrets
```
Download the private key sealed-secrets-ibm-demo-key.yaml used to seal any secret contained in this demonstration and apply it to the cluster. In our case, we have included a demo IBM Entitlement Key within the GitOps GitHub repository so that we are able to pull down IBM Software.
```
oc apply -f sealed-secrets-ibm-demo-key.yaml
```

Delete the pod

oc delete pod -n sealed-secrets -l app.kubernetes.io/name=sealed-secrets

IMPORTANT WARNING: DO NOT CHECK THE FILE INTO GIT

The private key MUST NOT be checked into GitHub under any circumstances. Please, remove the private key from your workstation to avoid any issues.
```
rm sealed-secrets-ibm-demo-key.yaml
```

For Cloud Pak to consume the entitlement key, restart the Platform Navigator pods

oc delete pod -n tools -l app.kubernetes.io/name=ibm-integration-platform-navigator

3 - RedHat OpenShift GitOps Operator¶

Clone the following GitHub repository that contains the GitOps structure that the Cloud Native Toolkit GitOps Framework understands.
```
git clone https://github.com/cloud-native-toolkit-demos/multi-tenancy-gitops-cp4d.git
```
Change directory into multi-tenancy-gitops-cp4d.
```
cd multi-tenancy-gitops-cp4d
```
Install the RedHat OpenShift GitOps operator on your RedHat OpenShift cluster and wait for it to be available:
- If your RedHat OpenShift cluster version is 4.6
```
oc apply -f setup/ocp46/
while ! kubectl wait --for=condition=Established crd applications.argoproj.io; do sleep 30; done
```
- If your RedHat OpenShift cluster version is 4.7
```
oc apply -f setup/ocp47/
while ! kubectl wait --for=condition=Established crd applications.argoproj.io; do sleep 30; done
```
Once the above command returns, you can open your RedHat OpenShift Web Console and check out that the RedHat OpenShift GitOps operator has been successfully installed in the openshift-gitops project.

As you can see in the image, the RedHat OpenShift GitOps operator also installs the RedHat OpenShift Pipelines operator and ArgoCD (which will be that GitOps tool that synchronizes the Infrastructure/Configuration as Code we have stored in GitHub with the state of the RedHat OpenShift cluster).

Important

The RedHat OpenShift Pipelines operator gets installed by the RedHat OpenShift GitOps Subscription only for RedHat OpenShift version 4.6. If your RedHat OpenShift cluster is version 4.7, you will need to install the RedHat OpenShift Pipelines operator as part of the GitOps process explained in this section. For getting such RedHat OpenShift Pipelines operator installed, you would need to specify that in the kustomize.yaml file for the services layer here.

Once ArgoCD is deployed, get the admin password

If your RedHat OpenShift cluster version is 4.6

oc extract secrets/argocd-cluster-cluster --keys=admin.password -n openshift-gitops --to=-

If your RedHat OpenShift cluster version is 4.7

oc extract secrets/openshift-gitops-cluster --keys=admin.password -n openshift-gitops --to=-

Open the ArgoCD web console by clicking on the ArgoCD console link you can see at the top of your RedHat OpenShift web console and log in.
Once you login, you should see that your ArgoCD web console is empty as we have not deployed any Argo Application yet.

4 - IBM Cloud Pak for Data¶

Before deploying an instance of IBM Cloud Pak for Data on the RedHat OpenShift cluster, we need to provide the storage node's IP addresses to the configuration of the OpenShift Data Foundation operator. You can provide custom configuration to the OpenShift Data Foundation operator in the ibm-odf.yaml file you can find in this IBM Cloud Pak for Data GitOps GitHub repository that was cloned at the beginning under multi-tenancy-gitops-process-mining/0-bootstrap/argocd/single-cluster/1-infra/argocd. Edit that file and provide your storage node's IP addresses to the worker_ip_X configuration parameters for the OpenShift Data Foundation operator.
Install the ArgoCD Bootstrap Application
```
oc apply -n openshift-gitops -f 0-bootstrap/argocd/bootstrap.yaml
```
This ArgoCD Bootstrap Application will bootstrap the deployment of IBM Process Mining based on the configuration you have defined in the GitOps GitHub repository we cloned earlier. You can see that we integrate Kustomize for configuration management in the GitOps approach.

As soon as you create this ArgoCD Bootstrap Application, the rest of the ArgoCD Applications and the respective RedHat Openshift resources these manage start to get created as a result of the synchronization process the GitOps approach is based on. You can see these ArgoCD Applications being created in the ArgoCD web console.
If you go to the Operators --> Installed Operators section of your RedHat OpenShift cluster web console and select the openshift-storage project in the Project drop down list at the top, you will see that the OpenShift Container Storage operator (which has been recently renamed to OpenShift Data Foundation) is being installed.
If you go to the Workloads --> Pods section of your RedHat OpenShift cluster web console you should see pods being created as a result of the OpenShift Container Storage operator being told to create an OpenShift Container Storage Cluster.
After some time, you should see the OpenShift Container Storage operator successfully installed

and the following new Storage Classes available on the Storage --> Storage Classes section of your RedHat OpenShift cluster web console

that will be used by the IBM Process Mining operator to create an IBM Process Mining instance.
If you go again to the Operators --> Installed Operators section of your RedHat OpenShift cluster web console and select the openshift-operators project in the Project drop down list at the top, you should see that the IBM Process Mining operator has been successfully installed as well as the IBM Automation Foundation Core and IBM Cloud Pak foundational services operators it depends on.
If you go to the Operators --> Installed Operators section of your RedHat OpenShift cluster web console and select the prod project in the Project drop down list at the top, since in our IBM Process Mining GitOps process we have configured the IBM Process Mining instance to be deployed in the prod project, and click on the IBM Process Mining operator and then on the Process Mining tab, you should see the IBM Process Mining instance and that this is Running and Ready.
If you go back to the ArgoCD web console, you should see all of the Argo Application in green.

5 - IBM Cloud Pak for Data UI¶

Let's make sure that our IBM Cloud Pak for Data instance is up and running. Do that by logging into the IBM Cloud Pak for Data user interface.

Obtain IBM Cloud Pak for Data console URL by executing

echo https://`oc -n cloudpak get ZenService lite-cr -o jsonpath="{.status.url}{'\n'}"`

Open the URL in a browser and you will be presented with the IBM Cloud Pak for Data user interface login option. Enter admin as username.

Obtain admin password by executing

oc -n cloudpak extract secret/admin-user-details --keys=initial_admin_password --to=-

Log into the IBM Cloud Pak for Data UI using the password from previous step.
Click on the navigation menu icon on the top left corner. Click on Services menu option to expand it, then select Services catalog.
The various services installed with IBM Cloud Pak for Data will be displayed.

That is it to get a working instance of IBM Cloud Pak for Data.