Structure and Performance of Decision Support Algorithms on Active Disks

Growth and usage trends for large decision support databases indicate that there is a need for architectures that scale the processing power as the dataset grows. These trends indicate that the processing demand for large decision support databases is growing faster than the improvement in performance of commodity processors. To meet this need, several researchers have recently proposed Active Disk/IDISK architectures which integrate substantial processing power and memory into disk units. In this paper, we examine the utility of Active Disks for decision support databases. We try to answer the following questions. First, is it possible to restructure algorithms for common decision support tasks to utilize Active Disks? Second, how does the performance of Active Disks compare with that of traditional servers for these tasks? Finally, how would Active Disks be integrated into the software architecture of decision support databases? 1 Introduction Growth and usage trends for large decis..