The OLAP Council is now sponsoring the development of an analytical processing benchmark, calling it the APB1. The benchmark is intended to simulate a real-life online analytical processing application and will measure the performance of analytical processing operations against databases, rather than the speed of the analytical processing server per se. Measured operations will include the bulk loading of data from internal or external data sources, incremental loading of data from operational systems; aggregation of input level data along hierarchies; calculation of new data based on business models; time series analysis; complex queries; drilldown hierarchies; ad hoc queries and multiple online sessions. Benchmarking will be carried out by independent auditors certified by the OLAP Council. Disclosure items will include an audit report, the database schema, code required for data loading plus any precalculations, code for the execution of each query, the number of users, the size of the benchmark data files, database size, the database server software, database tuning parameters and hardware and operating system parameters. The OLAP Council was due to release the benchmark for public comment last week. It has not worked out a pricing scheme as yet. In the meantime, the OLAP Council’s multi-dimensional data access application programming interface will be released into the public domain in the summer (CI No 2,592). The specification will supposedly lead to the creation of standard technologies for connecting front-end access software with online analytical processing data sources, according to the company.
