Selecting the Right Data Distribution Scheme for a Survivable Storage System (CMU-CS-01-120)

Survivable storage system design has become a popular research topic. This paper tackles the difficult problem of reasoning about the engineering trade-offs inherent in data distribution scheme selection. The choice of an encoding algorithm and its parameters positions a system at a particular point in a complex trade-off space between performance, availability, and security. We demonstrate that no choice is right for all systems, and we present an approach to codifying and visualizing this trade-off space. Using this approach, we explore the sensitivity of the space to system characteristics, workload, and desired levels of security and availability. This research is part of the PASIS project at Carnegie Mellon University.