Abstract: Parallel disk subsystems have been proposed as one way to close the gap between processor and disk speeds. In a previous paper we showed that prefetching and caching have the potential to deliver the performance benefits of parallel file systems to parallel applications. In this paper we describe experiments with practical prefetching policies, and show that prefetching can be implemented efficiently even for the more complex parallel file access patterns. We test these policies across a range of architectural parameters.
Keywords: parallel computing, file system, parallel-IO
Copyright © 1991 by IEEE.The copy made available here is the authors' version; for a definitive copy see the publisher's version described above.
See also earlier version kotz:thesis.
See also later version kotz:jpractical.