Performance evaluation of disk scheduling algorithms in the presence of bad sectors

Disk scheduling algorithms have long been a topic of study in computer science. Many researchers studied the performance of the disk scheduling algorithms. However, perhaps due to the difficulty of implementation, those early works focused solely on exploring the basic ideas and comparing the performance of these algorithms. No one studied the effect of bad sectors in their performance. In this paper, the performance of the disk scheduling algorithms in the presence of bad sectors is studied. The mapping of bad sectors to spare sectors is considered. We use simulation to do the performance evaluation. Simulation results show that all disk scheduling algorithms are same when there is a high percentage of bad sectors and when the spare sectors are distributed at the end of the disk. It is shown also that with a high percentage of bad sectors, the disk scheduling algorithms perform better if the distribution of spare sectors is at the end of the disk. With a low percentage of bad sectors, the scheduling algorithms perform better if the distribution of spare sectors is within the entire disk. It is shown also that all scheduling algorithms are more sensitive to bad sectors in heavily loaded systems.

[1]  Christopher Stewart,et al.  Realism and simplicity: disk simulation for instructional OS performance evaluation , 2006, SIGCSE '06.

[2]  Neil C. Wilhelm An anomaly in disk scheduling: a comparison of FCFS and SSTF seek scheduling using an empirical model for disk accesses , 1976, CACM.

[3]  Edward G. Coffman,et al.  Analysis of Scanning Policies for Reducing Disk Seek Times , 1972, SIAM J. Comput..

[4]  Michael A. Bender,et al.  New algorithms for the disk scheduling problem , 1996, Proceedings of 37th Conference on Foundations of Computer Science.

[5]  Alma Riska,et al.  It's not fair - evaluating efficient disk scheduling , 2003, 11th IEEE/ACM International Symposium on Modeling, Analysis and Simulation of Computer Telecommunications Systems, 2003. MASCOTS 2003..

[6]  Prashant J. Shenoy,et al.  Cello: A Disk Scheduling Framework for Next Generation Operating Systems* , 1998, SIGMETRICS '98/PERFORMANCE '98.

[7]  Yale N. Patt,et al.  Scheduling algorithms for modern disk drives , 1994, SIGMETRICS 1994.

[8]  Steven Robbins,et al.  A disk head scheduling simulator , 2004, SIGCSE '04.

[9]  Margo I. Seltzer,et al.  Disk Scheduling Revisited , 1990 .

[10]  Chang Liu,et al.  Disk scheduling policies with lookahead , 2002, PERV.

[11]  Steven Robbins A disk head scheduling simulator , 2004 .

[12]  Micha Hofri Disk scheduling: FCFS vs.SSTF revisited , 1980, CACM.

[13]  Prashant J. Shenoy,et al.  Cello: A Disk Scheduling Framework for Bext Generation Operating Systems , 1998, SIGMETRICS.

[14]  Darrell D. E. Long Proceedings of the Conference on File and Storage Technologies , 2002 .

[15]  Andrea C. Arpaci-Dusseau,et al.  USENIX Annual Technical ConferenceUSENIX Association 297 Robust , Portable I / O Scheduling with the Disk Mimic , 2003 .

[16]  Gregory R. Ganger,et al.  Towards higher disk head utilization: extracting free bandwidth from busy disk drives , 2000, OSDI.

[17]  Jason Nieh,et al.  Experiences teaching operating systems using virtual platforms and Linux , 2005, OPSR.

[18]  Jason Nieh,et al.  Experiences teaching operating systems using virtual platforms and linux , 2005 .