Introducing Red Hat® Marketplace, an open cloud marketplace that makes it easier to discover and access certified software for container-based environments in public clouds and on-prem. With automated deployment, software is immediately available to deploy on any Red Hat® OpenShift® cluster, providing a fast, integrated experience.
Check this out to know more about Red Hat Marketplace.
How to install, setup & get started using the Findability Platform Predict Plus operator on Red Hat OpenShift cluster
In this tutorial, we demonstrate how to install, setup and get started using the Findability Platform Predict Plus (FPPredict Plus) operator from Red Hat Marketplace (RHM). Every operator on Red Hat Marketplace has different setup for installation and configuration and the focus of this tutorial is Findability Platform Predict Plus operator.
The advantages of using RHM operators are per below.
-
Software for any cloud:- Enterprise software for container-based environments in public clouds and on-prem -
Automated deployment:- Fast, integrated experience with instant availability on Red Hat® OpenShift® clusters -
Fully supported:- Free, continuous support for products purchased through Red Hat Marketplace
FP-Predict +™, an Automated, self learning, and Multi Modeling Artificial Intelligence (AI) that handles Discrete Target variable, Continuous Target variable and Time series data with no need for coding. This operator will be helpful to solve some of the usecases under AI using different methods like Regression, Classification, Timeseries etc. This operator will be applicable to developers, data scientists, architects who wants to solve different usecases under machine learning and AI.
- For all operators being installed from RHM, OpenShift cluster version 4.3 or higher is mandatory. Please set up Classic cluster using the instructions from below URL.
- We need to have an account created at Red Hat Marketplace by following this link.
It would take about an hour to complete the tutorial.
Follow the steps below to launch the cluster console which is also called RedHat OpenShift Container Platform.
Login to IBM Cloud Account and navigate to Dashboard
Click on Clusters and select the cluster which you have created under prerequisites. In this tutorial, cluster name is cp-rhm-poc.
After you launch the cluster, click on OpenShift web console on the top right hand side.
We can see the RedHat OpenShift Container Platform (Web Console). Click on question mark icon on the top right hand side and select Command Line Tools.
Navigate to the section oc - OpenShift Command Line Interface (CLI) and download the respective oc binary onto your local system. This is needed to manage OpenShift projects from a terminal and is further extended to natively support OpenShift Container Platform features.
We are all set to proceed to next step which is to register the OpenShift cluster on Red Hat Marketplace platform. This is mandatory to install any operators from RedHat Marketplace platform using the OpenShift cluster.
Sign up and login to RHM portal at Link and click on workspace and then click on cluster. We need to add our new OpenShift cluster and register it on RHM platform.
Update the cluster name, generate the pull secret as per the instructions and save it.
Copy the curl command which starts with curl -sL https and append the pull secret towards the end. The entire script should be handy to be used in next step.
We need to start the cluster first to register it. Open a command prompt and type oc login, update the username and password which are used for accessing the cluster and hit enter.
The cluster is up and running at this point. We need to run the entire script on command prompt which is copied from previous step and hit enter or return in Mac. It will take a couple of mins and we can see that we have successfully registered the cluster on RedHat Marketplace portal.
We need to create a project to be used and managed from command line. Click on Create Project and give any name of your choice. We are using the name findability-project for this tutorial.
Sign in to Red Hat Marketplace, navigate to the search bar and select Findability Platform Predict Plus.
Choose the Free trial option in the next page and select the instance.
Note :- This is a trial instance with the following constraints per below.
- Up to 100K rows in training data and up to 50K rows in prediction data
- Up to 500 variables, features and columns
- 30 days free trial
Navigate to Workspace option at the top and click on My software option on the left pane.
Choose the operator FP-Predict+ by clicking on it and we will be directed to the next page where we need to choose Install operator from Operators menu.
Check the box under Target clusters and select the Name and choose the project name which was created in previous step under Namespace Scope option. We can leave the other default options as is and click Install. We are going ahead with Automatic method of installation, there is also an option do install the operator manually.
After a couple of minutes, the operator gets installed on the cluster. We should see the Status as Up to date. We can launch the web console by clicking on the option next to Cluster Console
In the Web console, we can verify by clicking on Installed Operators under Operators option in the left navigation pane and see that the operator is successfully installed with status showing as Succeeded.
We need to create persistent volume in the name of fp-predict-plus-pv to be bound to persistent volume claims. This step is necessary to enable storage capabilities for the operator to manage datasets. Click on persistent volume under Storage, update the name as fp-predict-plus-pv and storage capacity as 20 GB and hit Create. If we need more storage, we can increase it in the YAML file and create the persistent volume accordingly.
We need to create a persistent volume claim (storage) to use the storage created in previous step and bind it to the instance of this operator. Click on storage under Web Console and select Persistent Volume Claims.
The next step is to create persistent volume claim. We can select Storage Class from Gold, Silver, Bronze which was created in previous step, give the name as fp-predict-plus-pvc, select single user access & assign the storage size as 20 GB.
After the persistent volume claim (PVC) is created, it needs to be bound with the persistent volume (PV) which was created in earlier step. We should see the status as Bound per below after couple of minutes.
Click on Installed operators under Operators and click on FP Predict Plus Operator to get the options like Overview, YAML, Subscription, Events, FP-Predict-Plus. Click on YAML and update the name as fp-predict-plus-pvc per below, useExisting as false (useExisting would be true if we want to use the same storage for another instance), volumeName as fp-predict-plus-pv, routerCanonicalHostname would be the web console URL and hit Save. routerCanonicalHostname would start with the cluster name, cluster id till appdomain.cloud. The initial part in the URL - https://console-openshift-console should be removed while updating routerCanonicalHostname.
The next step is to proceed towards FP-Predict-Plus option and click on Create FPpredictplus instance. Edit the YAML file and give a name for the instance and click Save.
We are all set to launch the instance. How do we do it? We need to click on Networking and select Routes and then click on the URL which is under the location to launch the instance.
We can login using the default credentials as per the configuration file, accept the end user license agreement and will be directed to the next page per below. Click on Download and share the file with the Findability Sciences support team. The support team will send the license file (with 30 days validity) and needs to be uploaded using the Upload File option. We are all set to access the instance of FPPredict Plus.
Note :- Trial License key is provided only once by the support team. Please do not uninstall the instance of FP Predict Plus (in the web console or in Red Hat Marketplace portal) as re-installation of the instance would require a new trial license key.
After uploading the license key, we can bring up the FP Predict+ instance per below. The navigation pane on the left has four options namely Dashboard, Analytics, Dataset Management & License Information. We are good to go from here.
Note :- We may have to repeat all the steps if there is a version upgrade as some of the components from setup does not support product upgrades.
With this, we have come to the end of this tutorial to understand all the aspects of installation, configuration, setup and more to get started using the FPPredict Plus operator from Red Hat Marketplace on OpenShift cluster to solve usecases under AI. Red Hat Marketplace is a one stop platform for exploring multiple operators which will help to solve some of the complex usecases under different domains.