Data warehouses allow long term information about an enterprise to be recorded, summarized and presented.
A data warehouse is a passive observer object that takes no part in business processes, and is not part of the business model.
The axes of a multidimensional data warehouse are not arbitrary, but represent real aspects of the business.
Axes should represent the purpose, process, resource and organization aspects.
The summary hierarchies on each of these axes should parallel the fractal structures in the business model.
Roll up and drill down to zoom from summary to detail information is therefore based on the structure of the business, so is meaningful to management and other users.
Note that a financial ledger is a specialized form of data warehouse, where each general ledger account segment represents an axis.
Data warehouses and ledgers are objects and are efficiently implemented in Java™.
They are updated in real time by sending (asynchronous) messages containing the co-ordinates and value of the target warehouse cell (or ledger account) from business process objects.