Summary: Exploiting Gray­Box Knowledge of Buffer­Cache Management
Nathan C. Burnett, John Bent, Andrea C. Arpaci­Dusseau, and Remzi H. Arpaci­Dusseau
Department of Computer Sciences, University of Wisconsin--Madison
fncb, johnbent, dusseau, remzig@cs.wisc.edu
Abstract
The buffer­cache replacement policy of the OS can
have a significant impact on the performance of I/O­
intensive applications. In this paper, we introduce a
simple fingerprinting tool, Dust, which uncovers the re­
placement policy of the OS. Specifically, we are able to
identify how initial access order, recency of access, fre­
quency of access, and long­term history are used to de­
termine which blocks are replaced from the buffer cache.
We show that our fingerprinting tool can identify popu­
lar replacement policies described in the literature (e.g.,
FIFO, LRU, LFU, Clock, Random, Segmented FIFO,
2Q, and LRU­K) as well as those found in current sys­
tems (e.g., NetBSD, Linux, and Solaris).
We demonstrate the usefulness of fingerprinting the
cache replacement policy by modifying a web server to

Source: Arpaci-Dusseau, Remzi - Department of Computer Sciences, Department of Computer Sciences, University of Wisconsin at Madison
Liblit, Ben - Department of Computer Sciences, University of Wisconsin at Madison

Collections: Computer Technologies and Information Sciences