During SQL Adria conference, Namik Hrle (IBM Fellow) revealed some features that are likely to be delivered onto the DB2 Accelerator technology, hopefully soon. The DB2 Analytics Accelerator (formerly called IDAA) is an appliance that helps execute Business Intelligence (BI) queries much faster. DB2 for z/OS is built on an architecture that is optimized for transactional queries, but admittedly falls short when running complex analytics that nowadays business demands. The DB2 Analytics Accelerator is the solution that IBM is pursuing to provide high-performance BI queries against DB2 data (almost) transparently.
Namik used an interesting analogy, explaining that DB2 for z/OS with the DB2 Analytics Accelerator vision is to be like an hybrid car, that can decide to use either its petrol engine or its electrical engine, transparently to the driver (the end-user). Namik also indicated that the IBM lab is considering to provide the ability to:
* Create a table within the accelerator (that would reside solely in there, but would be accessible transparently through DB2)
* Provide a method to load non-DB2 data directly in the DB2 Analytics Accelerator (such as loading a flat file)
These 2 features would open the door to process complex BI queries, joining business data residing in DB2, and many other data sources, structures or unstructured …