Performance and safety considerations. The concurrency control methods of most DBMS do not react well to a mix of short update transactions (as in OLTP) and OLAP queries that typically search a large portion of the database. Moreover, the OLTP systems are often critical for the operation of the organization and must not be in danger of corruption by other applications.
Logical interpretability problems. Inspired by the success of spreadsheet techniques, OLAP users tend to think in terms of highly structured multidimensional data models, whereas information sources offer at best relational, often just semistructured data models or even flat files.
Temporal and granularity mismatch. OLTP systems focus on current operational support in great detail, whereas OLAP often considers historical developments in a somewhat lesser detail.
KeywordsUnify Modeling Language Data Warehouse Quality Function Deployment Enterprise Model Quality Goal
Unable to display preview. Download preview PDF.