My company is currently starting an enterprise data warehouse project with an end result of ultimately feeding all of the back-end systems including our General Ledger system. Is this the logical conclusion of an enterprise data warehouse? I am confused on the separation between ERP and a data warehouse.
Joe Oates’ Answer: First, an enterprise resource planning (ERP) system is an operational system. An ERP system is generally a commercial software package that integrates business functions and data across unit and division lines including sales, human resources, financial and operations. Implementing a full ERP system is a huge undertaking for most organizations and can even soak up the entire IT budget for up to several years. While most ERP vendors advertize that their ERP system covers all aspects of a business, including BI, there are still large areas that the software does not cover, and many or most ERP customers have a BI function that is different from the ERP vendor’s solution. Many organizations only implement the financials portion of the ERP suite.
A data warehouse is an analytical tool. Unlike ERP or other operational systems, the data warehouse is not updated every time an operational transaction takes place. The data warehouse is generally loaded on a daily, weekly or monthly basis with data from operational systems that is extracted and transformed. While many organizations have found that the data warehouse is an ideal place to keep the "single version of the truth" for data items such as customer information, this information is created or updated in a source system, transformed and loaded into the data warehouse, usually once a day (or night). It is only then that the customer or other information can be transformed into the format of other operational systems and sent from the data warehouse to other systems.
Something like the above can be used to feed operational systems and GL for certain types of information, but I am not sure that feeding all information to the GL from the data warehouse is a good idea. Having developed more than one accounting system, I know that end of period closings have a set number of business days to have the GL closed and that a lot of additions and deletions happen in the GL system that are not reflected anywhere else. I would have to understand your individual situation before recommending that all back-end systems be fed from the data warehouse. It is an intriguing idea, and might be a terrific solution, but I would have to study the situation carefully.
Clay Rehm’s Answer: I am confused as you are! The data warehouse should not be a source for any other systems, especially OLTP-type systems. If the goal of the enterprise data warehouse is to have an integrated store of current information, then you should really consider developing an operational data store (ODS). An ODS is meant as an integration point and typically feeds into data warehouses.
I would recommend you read the book, Building the Operational Data Store, by Inmon, Imhoff and Battas (ISBN 0-471-12822-8) for more detailed information.
The logical conclusion of a data warehouse is to support the decision-making process.
Les Barbusinskis Answer: If your company uses a completely integrated ERP system from a single vendor, with no third-party or custom-developed subsystems, there would be no reason to pipe data into your back-end systems through a data warehouse. However, if your application environment is heterogeneous, routing your back-end feed through the data warehouse provides the following benefits:
The question really boils down to one of cost vs. benefit.
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