Skip to main content

Accelerating Data Transformation with IBM DB2 Analytics Accelerator for z/OS

An IBM Redbooks publication


Published on 10 December 2015, updated 11 December 2015

  1. .EPUB (4.9 MB)
  2. .PDF (10.5 MB)

Apple BooksGoogle Play Books

Share this page:   

ISBN-10: 0738441198
ISBN-13: 9780738441191
IBM Form #: SG24-8314-00

Authors: Ute Baumbach, Patric Becker, Uwe Denneler, Eberhard Hechler, Wolfgang Hengstler, Steffen Knoll, Frank Neumann, Guenter Georg Schoellmann, Khadija Souissi and Timm Zimmermann

    menu icon


    Transforming data from operational data models to purpose-oriented data structures has been commonplace for the last decades. Data transformations are heavily used in all types of industries to provide information to various users at different levels. Depending on individual needs, the transformed data is stored in various different systems.

    Sending operational data to other systems for further processing is then required, and introduces much complexity to an existing information technology (IT) infrastructure. Although maintenance of additional hardware and software is one component, potential inconsistencies and individually managed refresh cycles are others.

    For decades, there was no simple and efficient way to perform data transformations on the source system of operational data. With IBM® DB2® Analytics Accelerator, DB2 for z/OS is now in a unique position to complete these transformations in an efficient and well-performing way. DB2 for z/OS completes these while connecting to the same platform as for operational transactions, helping you to minimize your efforts to manage existing IT infrastructure.

    Real-time analytics on incoming operational transactions is another demand. Creating a comprehensive scoring model to detect specific patterns inside your data can easily require multiple iterations and multiple hours to complete. By enabling a first set of analytical functionality in DB2 Analytics Accelerator, those dedicated mining algorithms can now be run on an accelerator to efficiently perform these modeling tasks.

    Given the speed of query processing on an accelerator, these modeling tasks can now be performed much quicker compared to traditional relational database management systems. This speed enables you to keep your scoring algorithms more up-to-date, and ultimately adapt more quickly to constantly changing customer behaviors.

    This IBM Redbooks® publication describes the new table type that is introduced with DB2 Analytics Accelerator V4.1 PTF5 that enables more efficient data transformations. These tables are called accelerator-only tables, and can exist on an accelerator only.

    The tables benefit from the accelerator performance characteristics, while maintaining access through existing DB2 for z/OS application programming interfaces (APIs). Additionally, we describe the newly introduced analytical capabilities with DB2 Analytics Accelerator V5.1, putting you in the position to efficiently perform data modeling for online analytical requirements in your DB2 for z/OS environment.

    This book is intended for technical decision-makers who want to get a broad understanding about the analytical capabilities and accelerator-only tables of DB2 Analytics Accelerator. In addition, you learn about how these capabilities can be used to accelerate in-database transformations and in-database analytics in various environments and scenarios, including the following scenarios:

    • Multi-step processing and reporting in IBM DB2 Query Management Facility™, IBM Campaign, or Microstrategy environments
    • In-database transformations using IBM InfoSphere® DataStage®
    • Ad hoc data analysis for data scientists
    • In-database analytics using IBM SPSS® Modeler

    Table of Contents

    Chapter 1. Analytics on an IBM z Systems environment

    Chapter 2. Accelerator-only tables

    Chapter 3. Use cases that are enabled by accelerator-only tables and in-database analytics

    Chapter 4. Multistep reporting

    Chapter 5. Using IBM DB2 QMF to store query results and import tables

    Chapter 6. Accelerating IBM Campaign processing

    Chapter 7. In-database transformations

    Chapter 8. Accelerator-only tables supporting data scientists' ad hoc analysis

    Chapter 9. Integrating more data sources and archiving for analytics

    Chapter 10. In-database analytics

    Appendix A. Description of IBM z Systems environment used for this publication


    Others who read this also read