If you have young children like I do, you've read the story of the "Little Red Hen." In case you haven't recently read this page turner, a hen in the barnyard wants to make some bread, has trouble motivating the other animals to help, makes it herself and then has lots of hungry friends. Since the most valuable things we learn in life are often learned in kindergarten, I suggest the Little Red Hen has deeper meaning. In her quest to provide bread for the barnyard, she encounters passive resistance--"Everyone wants to eat the bread, but no one eats to bake the bread!" What the Little Red Hen didn't realize was that like most of us, she is living in an outcome-based barnyard, where results matter most.
Barnyard animals don't have time to create bread from scratch, they just want to be fed. The Little Red Hen would be better served to get some helpers and explore using bread baking mix or outsource the bread production altogether. In data warehousing, helpers cut development time and maximize results by focusing knowledge and effort on the value-added contributions.
Helpers for a data warehouse consist of a hierarchy of products, tools, utilities, templates and pre-built application frameworks. There are different helpers for each layer of the data warehouse: transformation rules, meta data rules and the user interface (see last month's article for detail on the layers).
Early data warehousing transformation logic and data cleansing rules were coded in COBOL or by using data extraction utilities. This approach can embed business rules in the transformation logic, making it hard for end users to validate the rules. In addition, generic rules are not easily shared with new data source transformation needs. More recently, tools specifically designed for transformation have replaced most of the proprietary approaches. Tools from Informatica, Sagent and Ardent allow the developer to abstract the business and cleansing rules into a reusable layer, generating the transformation logic and data output to multiple target databases. These transformation helpers simplify the initial population of the data warehouse and reduce the maintenance down the line.
The meta data layer has fewer choices for helpers than the transformation layer. Most managed query environments (MQE) such as those from companies such as Business Objects, Cognos or Brio offer a meta data repository for their application environment. These repositories can also be accessed by other vendor products to some degree and greatly increase efficiency in development. The only drawback is that locking into a proprietary meta data repository may limit other choices down the line. This may or may not be an issue, depending on the design and development of your data warehouse. In any case, using an available meta data repository can free you from the tasks of defining, constructing and managing the repository logic.
The most recent arrivals on the helper scene are industry specific templates and "quick start" applications. Templates can take the form of rule packages, data models or pre-built reports and queries. The use of templates can significantly shorten the development time for core subject areas in the data warehouse. "Quick start" applications carry these components one step further and deliver templates wrapped together with an "in-the-box" working application solution. Just map your source data into the transformation layer, flip the switch to populate the initial data mart and start producing canned queries, views and reports. If the conceptual architecture of the "quick start" application is sound and flexible for enhancement, these applications offer a quick return on investment by enabling organizations to go from planning to implementation in as few as three months.
Using data warehouse helpers can reduce risk, increase value and speed deliverables to the organization or barnyard, as the case may be. Don't be disheartened when no one wants to make the bread or help build the data warehouse. The world is filled with users hungry for information. Bake the bread, create the data warehouse and they will come for a meal.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access