Table of contents

Creating an integrated database

You can optionally create one or more databases within IBM® Cloud Pak for Data. After you create an integrated database, you can create a connection to the database so that the contents of the database can be discovered, governed, and used for analytics.

Integrated databases can be used to:
  • Offload some of your data workload to the system where the governance and analysis occurs
  • Quickly provision databases if you don't have existing databases
  • Create a sandbox environment

You can create several different types of integrated databases in Cloud Pak for Data.

Important: The required resources for each database are the minimum requirements. You should size your database based on your expected workload.

Detailed requirements and use cases for integrated databases provides more in-depth information on planning for an integrated database.

IBM Db2 Advanced Enterprise Server Edition

Relational database that delivers advanced data management and analytics capabilities for transactional workloads.

  • Nodes:
    • 1 node for standard deployments
    • 3 nodes for HADR deployments
  • Supported storage:
    • hostPath storage (not supported on HADR deployments)
    • NFS storage
  • CPU and memory:

    Varies based on your expected workload

For more information, see IBM Db2 Advanced Enterprise Server Edition in Detailed requirements and use cases for integrated databases.

IBM Db2 Event Store

A data store capable of high speed ingest and deep, real-time analytics.

  • Nodes: 3
  • Supported storage:
    • IBM Cloud Object Storage
    • GlusterFS storage
  • Local storage (per node): 100 GB SSD with XFS format
  • CPU and memory:

    Varies based on your expected workload

For more information, see IBM Db2 Event Store in Detailed requirements and use cases for integrated databases.

IBM Db2 Warehouse

A high-performing analytic engine that combines in-memory processing with integrated database analytics. Runs on either a single node, for cost-efficiency, or on multiple nodes for better performance.

  • Nodes: 1, or 2-999
  • Supported storage:
    • hostPath storage
    • NFS storage
  • CPU and memory:

    Varies based on your expected workload

For more information, see IBM Db2 Warehouse in Detailed requirements and use cases for integrated databases.

MongoDB

MongoDB is a flexible and scalable database that is designed to store, query, and index documents.

  • Nodes: 1
  • Supported storage:
    • Local storage
    • NFS storage
  • CPU and memory:

    Varies based on your expected workload

For more information, see MongoDB in Detailed requirements and use cases for integrated databases.

PostgreSQL

PostgreSQL is a popular, open source SQL database that is extensible, ACID compliant, and supports high availability and backup and restore functions.

  • Nodes: 2 (minimum)
  • Supported storage:
    • NFS storage
    • Local storage
  • CPU and memory:

    1 core, 8GB (minimum)

For more information, see PostgreSQL in Detailed requirements and use cases for integrated databases.

Creating a database

To create an integrated database, you must complete the following tasks:
  1. Downloading the database installation package.
  2. Preparing your IBM Cloud Private cluster for a database.
  3. Configuring database storage.
  4. Creating a database deployment on the cluster.