A File System for Intelligent Disks

Randy Wang


Our latest work on xFS has focused on the use of intelligent I/O processors. In this talk, we explore the role of a programmable disk processor in improving file system performance. Existing file systems suffer from two common fundamental assumptions that handicap their designs: that a small disk write is intrinsically expensive and that reorganizing the disk layout is a heavy-weight operation. By dedicating part of the storage management to the intelligent embedded disk processor, we are faced with the opportunity of rethinking a new file system paradigm based on low-latency persistent writes and light-weight reorganization. This is possible because the IDISK file system is able to take advantage of its intimage knowledge of both of the underlying disk mechanisms and file system semantics. We have developed a number of analytical models which guide the construction of efficient IDISK allocation algorithms. Preliminary simulation results promise significant performance improvement.
Click here to go back to the Finale schedule.