DecisionBase is a tool for moving and transforming data targeted at the data-warehouse market. Clearly, it is aimed at large enterprises and it is typically sold with services to set it up. I worked with Neal Peckham, ace Platinum consultant, for about one day to get the product running.
Version 1.8 of Platinum DecisionBase from Platinum Technology is made up of three components: Repository, Mapper and InfoPump (a movement server). Together, these components should meet the data movement and transformation needs of almost anyone.
Let’s start with the Repository, which is Platinum’s standard repository for meta data. It is part of Platinum’s Open Enterprise Management System, Platinum’s foundation for building enterprise applications. The Repository includes scanners to examine different databases, copybooks, and other data sources, and it will populate the repository with meta data.
An added bonus is that the Repository includes support for several major enterprise resource planning implementations, including SAP and PeopleSoft. This means if you have SAP, the Repository understands it and lists the tables included with SAP with names in English instead of German. However, these stop at the table level.
The Repository also includes an HTML-based application which provides a highly extensible browser-driven interface for looking at meta data. The browser makes the data accessible to more people, who then have a better idea of what’s going on based on the data they see.
The Mapper is where the magic happens with DecisionBase. It provides a listing of all data in the Repository and provides a visual environment for mapping source data to targets. Once you have completed your transformation application, the Mapper generates the code that will actually run the transformation from the source data to the target.
DecisionBase’s programming language is robust. It supports scripts or data elements as large as 2GB, 40-plus data types, multibyte character sets and exact decimal numbers. In short, it’s built to deal with complex data movement and transformation issues.
I liked the Entity Relationship Diagram-like look of the Mapper, which provided a clear picture of sources, targets, transformations and the links among them. DecisionBase is easier to work with than Compuware’s FileAID C/S, which performs a similar function with a less intuitive interface.
The final piece of the puzzle is the movement server, which is Platinum’s InfoPump. InfoPump processes the data movement and transformation by executing the code generated in the Mapper. You won’t really notice the movement server unless you need to administer it. It runs jobs created in the Mapper and includes tools for managing processes and scheduling jobs.
Platinum offers InfoPump by itself, priced at about US$30,000 with some consulting. To use it, you will have to write the scripts that DecisionBase’s Mapper automatically generates. It would be nice to see the Mapper bundled without the repository for this price for smaller customers that need a robust tool that is easy to use but do not need an enterprise repository to store meta data.
There’s no question that DecisionBase is a great tool for moving and transforming data for those seeking to populate data warehouses from diverse data sources. Will you commit to the Platinum repository for your enterprise meta-data store, because for US$200,000 you’re unlikely to implement another repository? If the answer is yes, DecisionBase is the ultimate weapon.
Hammond is a freelance writer and a developer at View-mark, a Colorado-based new media design company.
