Cloud deployment environments

You can choose to deploy IBM Cloud Pak® for Data in the environment that suits your business needs.

Cloud Pak for Data can be deployed in various private cloud and public cloud environments.

Deployment environment Managed OpenShift® Self-managed OpenShift
On-premises ✓ ✓
IBM® Cloud ✓ ✓
Amazon Web Services (AWS) ✓ ✓
Microsoft Azure ✓ ✓
Google Cloud   ✓
Manual deployments
You can deploy Cloud Pak for Data on any supported cloud deployment environment by using the cli installation method. For more information, see Manual deployment options.
Automated deployments
You can deploy Cloud Pak for Data on select cloud deployment environments by using a marketplace offering. The marketplace offerings enable you to install Cloud Pak for Data quickly and easily but have several limitations. For more information, see Automated deployment options.
Important: If you plan to deploy a service that requires GPU, ensure that the deployment environment that you choose supports the required GPU. For example, the watsonx.ai service requires NVIDIA A100 or H100 GPUs with 80 GB RAM, or L40S GPUs with 48 GB RAM, which are not available on Red Hat OpenShift Service on AWS (ROSA). For more information about which services require GPU, see Hardware requirements.

Manual deployment options

With a manual installation of Cloud Pak for Data, you have complete control over how Cloud Pak for Data is installed on your cluster.

The following sections include additional guidance about:
  • How to install OpenShift on the environment
  • The recommended cluster architecture
  • How to install the Cloud Pak for Data platform and services on the environment
On-premises infrastructure

If you want to ensure that your environment is running securely behind your firewall, or you have an existing on-premises Red Hat OpenShift Container Platform cluster, you can deploy Cloud Pak for Data on your on-premises infrastructure.

Installing Red Hat OpenShift Container Platform
The method for installing Red Hat OpenShift Container Platform depends on whether you want a managed or self-managed OpenShift cluster:
OpenShift type Deployment options
Managed You can deploy a managed OpenShift cluster on your on-premises infrastructure with IBM Cloud Satellite. For details, see IBM Cloud Satellite in the IBM Cloud marketplace.
Self-managed You can deploy a self-managed OpenShift cluster on your on-premises infrastructure by following the Red Hat OpenShift Container Platform documentation:

Additional guidance on setting up OpenShift is available in the IBM Cloud Paks documentation.
Tip: If your license includes entitlement to Red Hat OpenShift Container Platform, you can apply the entitlement to your self-managed OpenShift cluster.
Recommended cluster architecture
Cloud Pak for Data is deployed on a multi-node cluster. Although you can deploy Cloud Pak for Data on a 3-node cluster for development or proof of concept environments, it is strongly recommended that you deploy your production environment on a larger, highly available cluster with multiple dedicated control plane and worker nodes. This configuration provides better performance, better cluster stability, and increased ease of scaling the cluster to support workload growth. The specific requirements for a production-level cluster are identified in Hardware requirements.

For more information, see Architecture for IBM Cloud Pak for Data.

Supported storage
Deployment environment Managed OpenShift Self-managed OpenShift
On-premises IBM Cloud Satellite supports the following storage options with managed OpenShift:
  • OpenShift Data Foundation
  • Portworx
 The following storage options are supported on bare metal and VMware infrastructure with self-managed OpenShift:
  • OpenShift Data Foundation
  • IBM Storage Fusion Data Foundation
  • IBM Storage Fusion Global Data Platform
  • IBM Storage Scale Container Native
  • Portworx
  • NFS
  • NetApp Trident
Prerequisites
Ensure that you review the following information:
Topic What to look for
Hardware requirements Based on the services that you plan to install, determine whether you have sufficient hardware and whether you need AVX instructions or dedicated nodes.
Software requirements Based on the services that you plan to install, determine whether you have the necessary external software and service dependencies.
Storage requirements The storage that you use determines the services that you can run. Ensure that the services that you want to run are compatible with the storage in your environment.
Installing Cloud Pak for Data and services
See Installing IBM Cloud Pak for Data.
IBM Cloud infrastructure

If you already use IBM Cloud to run business-critical applications, or if you don't want to set up and manage your own hardware, you can deploy Cloud Pak for Data on IBM Cloud.

Installing Red Hat OpenShift Container Platform

The method for installing Red Hat OpenShift Container Platform depends on whether you want a managed or self-managed OpenShift cluster:

OpenShift type Deployment options
Managed You can deploy a managed OpenShift cluster on IBM Cloud infrastructure from the IBM Cloud catalog. For details, see Red Hat OpenShift on IBM Cloud® in the IBM Cloud catalog.

Ensure that you select a supported version of Red Hat OpenShift.

You can deploy Cloud Pak for Data on Virtual Private Cloud (VPC) infrastructure. You can use either a single zone or multi zone deployment.
Self-managed You can deploy a self-managed OpenShift cluster on Virtual Private Cloud (VPC) IBM Cloud infrastructure by following the Red Hat OpenShift Container Platform documentation:

Additional guidance on setting up OpenShift is available in the IBM Cloud Paks documentation.
Tip: If your license includes entitlement to Red Hat OpenShift Container Platform, you can apply the entitlement to your self-managed OpenShift cluster.
Recommended cluster architecture
Cloud Pak for Data is deployed on a multi-node cluster. Although you can deploy Cloud Pak for Data on a 3-node cluster for development or proof of concept environments, it is strongly recommended that you deploy your production environment on a larger, highly available cluster with multiple dedicated control plane and worker nodes. This configuration provides better performance, better cluster stability, and increased ease of scaling the cluster to support workload growth. The specific requirements for a production-level cluster are identified in Hardware requirements.

For more information, see Architecture for IBM Cloud Pak for Data.

Supported storage
Deployment environment Managed OpenShift Self-managed OpenShift
IBM Cloud Red Hat OpenShift on IBM Cloud supports the following storage options on VPC infrastructure:
  • OpenShift Data Foundation
  • IBM Storage Fusion Data Foundation
  • Portworx
 The following storage options are supported on VPC IBM Cloud infrastructure with self-managed OpenShift:
  • OpenShift Data Foundation
  • IBM Storage Fusion Data Foundation
  • Portworx
  • NFS
Prerequisites
Ensure that you review the following information:
Topic What to look for
Hardware requirements Based on the services that you plan to install, determine whether you have sufficient hardware and whether you need AVX instructions or dedicated nodes.
Software requirements Based on the services that you plan to install, determine whether you have the necessary external software and service dependencies.
Storage requirements The storage that you use determines the services that you can run. Ensure that the services that you want to run are compatible with the storage in your environment.
Installing Cloud Pak for Data and services
See Installing IBM Cloud Pak for Data.
Amazon Web Services (AWS) infrastructure

If you already use AWS and you don't want to set up and manage your own hardware, you can deploy Cloud Pak for Data on AWS.

Installing Red Hat OpenShift Container Platform

The method for installing Red Hat OpenShift Container Platform depends on whether you want a managed or self-managed OpenShift cluster:

OpenShift type Deployment options
Managed
ROSA CLI
You can use the ROSA CLI to manually install ROSA. For details, see the Red Hat OpenShift Service on AWS (ROSA) documentation.
IBM Cloud Satellite
You can use IBM Cloud Satellite in the IBM Cloud marketplace to install Red Hat OpenShift Container Platform on AWS.
Self-managed
You can deploy a self-managed OpenShift cluster on AWS infrastructure by following the Red Hat OpenShift Container Platform documentation:

Additional guidance on setting up OpenShift is available in the IBM Cloud Paks documentation.
Tip: If your license includes entitlement to Red Hat OpenShift Container Platform, you can apply the entitlement to your self-managed OpenShift cluster.
Recommended cluster architecture
Cloud Pak for Data is deployed on a multi-node cluster. Although you can deploy Cloud Pak for Data on a 3-node cluster for development or proof of concept environments, it is strongly recommended that you deploy your production environment on a larger, highly available cluster with multiple dedicated control plane and worker nodes. This configuration provides better performance, better cluster stability, and increased ease of scaling the cluster to support workload growth. The specific requirements for a production-level cluster are identified in Hardware requirements.

For more information, see Architecture for IBM Cloud Pak for Data.

Supported storage
Deployment environment Managed OpenShift Self-managed OpenShift
Amazon Web Services (AWS) Red Hat OpenShift Service on AWS (ROSA) supports the following storage options:
  • IBM Storage Fusion Global Data Platform
  • Amazon Elastic Block Store (EBS)
  • Amazon Elastic File System (EFS)
  • NetApp Trident (includes Amazon FSx for NetApp ONTAP)
 The following storage options are supported on AWS infrastructure with self-managed OpenShift:
  • OpenShift Data Foundation
  • IBM Storage Fusion Data Foundation
  • Amazon Elastic Block Store (EBS)
  • Amazon Elastic File System (EFS)
  • Portworx
  • NFS
  • NetApp Trident (includes Amazon FSx for NetApp ONTAP)
Prerequisites
Ensure that you review the following information:
Topic What to look for
Hardware requirements Based on the services that you plan to install, determine whether you have sufficient hardware and whether you need AVX instructions or dedicated nodes.
Software requirements Based on the services that you plan to install, determine whether you have the necessary external software and service dependencies.
Storage requirements The storage that you use determines the services that you can run. Ensure that the services that you want to run are compatible with the storage in your environment.
Installing Cloud Pak for Data and services
See Installing IBM Cloud Pak for Data.
Microsoft Azure infrastructure

If you already use Microsoft Azure and you don't want to set up and manage your own hardware, you can deploy Cloud Pak for Data on Azure.

Installing Red Hat OpenShift Container Platform
OpenShift type Deployment options
Managed
You can use the az aro CLI to manually install Azure Red Hat OpenShift 4 (ARO). For details, see the Azure Red Hat OpenShift 4 documentation.
Self-managed
You can deploy a self-managed OpenShift cluster on Azure infrastructure by following the Red Hat OpenShift Container Platform documentation:

Additional guidance on setting up OpenShift is available in the IBM Cloud Paks documentation.
Tip: If your license includes entitlement to Red Hat OpenShift Container Platform, you can apply the entitlement to your self-managed OpenShift cluster.
Recommended cluster architecture
Cloud Pak for Data is deployed on a multi-node cluster. Although you can deploy Cloud Pak for Data on a 3-node cluster for development or proof of concept environments, it is strongly recommended that you deploy your production environment on a larger, highly available cluster with multiple dedicated control plane and worker nodes. This configuration provides better performance, better cluster stability, and increased ease of scaling the cluster to support workload growth. The specific requirements for a production-level cluster are identified in Hardware requirements.

For more information, see Architecture for IBM Cloud Pak for Data.

Supported storage
Deployment environment Managed OpenShift Self-managed OpenShift
Microsoft Azure Azure Red Hat OpenShift (ARO) supports the following storage options:
  • OpenShift Data Foundation
  • IBM Storage Fusion Data Foundation
  • Portworx
  • NFS
 The following storage options are supported on Microsoft Azure infrastructure with self-managed OpenShift:
  • OpenShift Data Foundation
  • IBM Storage Fusion Data Foundation
  • Portworx
  • NFS, specifically Microsoft Azure locally redundant Premium SSD storage
Prerequisites
Ensure that you review the following information:
Topic What to look for
Hardware requirements Based on the services that you plan to install, determine whether you have sufficient hardware and whether you need AVX instructions or dedicated nodes.
Software requirements Based on the services that you plan to install, determine whether you have the necessary external software and service dependencies.
Storage requirements The storage that you use determines the services that you can run. Ensure that the services that you want to run are compatible with the storage in your environment.
Installing Cloud Pak for Data and services
See Installing IBM Cloud Pak for Data.
Google Cloud infrastructure

If you already use Google Cloud and you don't want to set up and manage your own hardware, you can deploy Cloud Pak for Data on Google Cloud.

Installing Red Hat OpenShift Container Platform
OpenShift type Deployment options
Managed Managed OpenShift on Google Cloud infrastructure is not supported.
Self-managed
You can deploy a self-managed OpenShift cluster on Google Cloud infrastructure by following the Red Hat OpenShift Container Platform documentation:

Additional guidance on setting up OpenShift is available in the IBM Cloud Paks documentation.
Remember: If your license includes entitlement to Red Hat OpenShift Container Platform, you can apply the entitlement to your self-managed OpenShift cluster.
Recommended cluster architecture
Cloud Pak for Data is deployed on a multi-node cluster. Although you can deploy Cloud Pak for Data on a 3-node cluster for development or proof of concept environments, it is strongly recommended that you deploy your production environment on a larger, highly available cluster with multiple dedicated control plane and worker nodes. This configuration provides better performance, better cluster stability, and increased ease of scaling the cluster to support workload growth. The specific requirements for a production-level cluster are identified in Hardware requirements.

For more information, see Architecture for IBM Cloud Pak for Data.

Supported storage
Deployment environment Managed OpenShift Self-managed OpenShift
Google Cloud Managed OpenShift on Google Cloud is not supported. The following storage options are supported on Google Cloud infrastructure with self-managed OpenShift:
  • OpenShift Data Foundation
  • Portworx
  • NFS
Prerequisites
Ensure that you review the following information:
Topic What to look for
Hardware requirements Based on the services that you plan to install, determine whether you have sufficient hardware and whether you need AVX instructions or dedicated nodes.
Software requirements Based on the services that you plan to install, determine whether you have the necessary external software and service dependencies.
Storage requirements The storage that you use determines the services that you can run. Ensure that the services that you want to run are compatible with the storage in your environment.
Installing Cloud Pak for Data and services
See Installing IBM Cloud Pak for Data.

In addition to Cloud Pak for Data software, IBM offers IBM Cloud Pak for Data as a Service. IBM Cloud Pak for Data as a Service might be right for you if you already use IBM Cloud to run business-critical applications and you don't want to set up and manage your own deployment of Cloud Pak for Data. IBM Cloud Pak for Data as a Service differs from the Cloud Pak for Data software in several ways. For details, see the IBM Cloud Pak for Data as a Service documentation.

Automated deployment options

Depending on the environment , you might be able to install Cloud Pak for Data and some services from your cloud provider's marketplace.

The marketplace installation options are offered in addition to the default cli installation method.

The marketplace installations have several limitations:
  • Some marketplace installations automatically create a Red Hat OpenShift Container Platform cluster and set up storage for you. You cannot use the marketplace offering to install Cloud Pak for Data on an existing cluster.
  • Market place installations support a subset of the storage that Cloud Pak for Data supports.
  • Marketplace installations do not support installation from a private container registry. If you want to use a private container registry, you must use the cli installation method.
  • Marketplace installations support only installation. If you want to upgrade an existing installation, you must use the cli to upgrade.
  • Marketplace installations can be used to automatically install a subset of Cloud Pak for Data services. However, the services that are available depend on the cloud deployment environment. If you want to install additional services, you must use the cli installation method.
Important: The automated deployment options might not support the latest version of Cloud Pak for Data. The automated deployment options are updated after the Cloud Pak for Data software is released.

Each marketplace offering specifies the version of Cloud Pak for Data that it will install. If you want to install a different version of Cloud Pak for Data, you must use the cli installation method.

IBM Cloud infrastructure

You can install Cloud Pak for Data on Red Hat OpenShift on IBM Cloud from the IBM Cloud Catalog.

Red Hat OpenShift Container Platform
Review the following information to ensure that the marketplace offering supports the type of environment where you want to run Cloud Pak for Data.
Environment Details
Managed OpenShift Supported.
Self-managed OpenShift No marketplace offering is available for self-managed Red Hat OpenShift Container Platform on IBM Cloud.
Prerequisites
Before you can install Cloud Pak for Data from the IBM Cloud Catalog, you must:
  1. Install Red Hat OpenShift on IBM Cloud.

    For details, see Red Hat OpenShift on IBM Cloud in the IBM Cloud catalog.

    Ensure that you select a supported version of Red Hat OpenShift.

    You can deploy Cloud Pak for Data on Virtual Private Cloud (VPC) Gen2 single or multi zone infrastructure.

  2. Set up storage.

    This marketplace offering supports OpenShift Data Foundation storage.

Amazon Web Services (AWS) infrastructure

You can install Cloud Pak for Data on AWS from the AWS marketplace.

Red Hat OpenShift Container Platform
Review the following information to ensure that the marketplace offering supports the type of environment where you want to run Cloud Pak for Data.
Environment Details
Managed OpenShift Supported.
Important: Red Hat OpenShift Service on AWS (ROSA) is supported only for Data Fabric offerings.
Self-managed OpenShift Supported.
Important: Self-managed OpenShift is supported for BYOL offerings and cartridge offerings.
About this offering
The AWS marketplace offerings will complete the following tasks for you:
  • Set up a Red Hat OpenShift Container Platform cluster.
  • Set up storage.
    You can choose from the following options:
    • Red Hat OpenShift Data Foundation storage
    • Amazon Elastic File System (EFS) storage and Amazon Elastic Block Store (EBS)
    • Amazon Elastic File System (EFS) storage only
  • Install Cloud Pak for Data and selected services.
Prerequisites

Before you can install Cloud Pak for Data from the AWS marketplace, you must have the following information:

Microsoft Azure infrastructure

You can install Cloud Pak for Data on Azure from the Azure Marketplace.

Red Hat OpenShift Container Platform
Review the following information to ensure that the marketplace offering supports the type of environment where you want to run Cloud Pak for Data.
Environment Details
Managed OpenShift Supported.
Important: To install on Azure Red Hat OpenShift (ARO), choose an offering with ARO in the name.
Self-managed OpenShift Supported.
About this offering
The Azure marketplace offerings will complete the following tasks for you:
  • Set up a Red Hat OpenShift Container Platform cluster.
  • Set up Red Hat OpenShift Data Foundation storage.
  • Install Cloud Pak for Data and selected services.
Prerequisites

Before you can install Cloud Pak for Data from the Azure marketplace, you must have the following information: