As I noted in Part I, when attempting to develop a real-world performance test for BI tools, "the data matters." So, to that point, and to keep this as focused on what's important as necessary, a brief comment on the data requirements for the Business Intelligence Performance Index, or The BIPI.
• The content of the data matters.
• The “shape” of the data matters.
• The amount of data matters.
The data must be “real”, not randomly generated.
The data model must reflect an actual Business Intelligence model for analytical queries and reports.
The data size must, at a minimum, represent typical analytical databases, data warehouses and data marts. The BIPI specifies no less than 1 terabyte (TB) in overall raw data size, calculated at the file system level as a sum of the size of the load files, and no less than 1 billion records in at least one table in the model.
Stay tuned for The BIPI - Part III, The Execution
Post Comment