A study of I/O system organizations
Pages 308 - 317
Abstract
With the increasing processing speeds, it has become important to design powerful and efficient I/O systems. In this paper, we look at several design options in designing an I/O system and study their impact on the performance. Specifically, we use trace driven simulations to study a disk system with a nonvolatile cache. Some of the considered design parameters include the cache block size, the fetch size, the cache size and the disk access policy. We show that decoupling the fetch size and the cache block size results in significant performance improvements. A new write-back policy is presented that is shown to offer significant performance benefits. We show that optimal block size in a two-level memory hierarchy is dependent only on the latency, data rate product of the second level as previously conjectured. We also present results showing the effect of a split access operation of a disk read/write head.
References
[1]
G. M. Amdahl. Validity of the single processor approach to achieving large scale computing capabilities. Proc. AF{PS Spr, ng Joznt Comput. Conf., 30:483-485, April 1967.
[2]
J. Akella and D. P. Siewiorek. Modeling and measurement of the impact of input/output on system performance. Proc. of 18th Ann. Syrup. on Computer Arch., M~y 1991.
[3]
D. A. Patterson, G. Gibson, and R. H. Katz. A case for redundant arrays of inexpensive disks (RAID). A CM SIGMOD Conference, June 1988.
[4]
A. L. Narasimha Reddy and P. Banerjee. An evaluation of multiple-disk I/O systems. IEEE Trans. Comput., C-38, no. 12"1680-1690, Dec. 1989.
[5]
K. Salem and H. Garcia-Molina. Disk striping. Int. Co~tf. on Data gngznccrzT~g, pages 336-342, 1986.
[6]
M. Livny, S. Khoshafian, and H. Boral. Multi-disk management algorithnls. Proc. ACM SIGMET- RICS Conf., pages 69-77, May 1987.
[7]
M. Y. Kim. Synchronized disk interleaving. IEEE Y a,s. Co,l,u*., c-an,o. 11.978-988, Nov. 1986.
[8]
A. J. Smith. Disk cache-miss ratio analysis and design considerations. A CM Trans. on Comput. Systems, 3, no. 3161-203, Aug. 1985.
[9]
J. K. Ousterhout, H. DaCosta, D. Harrison, a. Kunze, M. Kupfer, and J. Tlmmpson. A tracedriven analysis of the unix 4.2 bsd file system. Proc. lOlh Syrup. on Operatzng System Pmnczples, pages 15-24, Dec. 1985.
[10]
J. Ousterhout and F. Douglis. Beating the I/O bottleneck" A case for log-structured file systems. Tech. Rep., Dept. of EECS, Univ. of California, Berkeley, Aug. 1988.
[11]
M. Rosenblum and a. K. Ousterhout. The LFS storage manager. Proc. of A CM Syrup. on Oper. Syst. Prznczples, 1991.
[12]
The SPEC benchmark suite. SPEC, Fremont, CA, Dec. 1990.
[13]
A. L. Narasimha Reddy. Reads and writes: when I/Os aren't quite the same. IBM Tech. Report: RJ 8033, March 1991.
[14]
A. a. Smith. Line (block) size choice for cpu cache memories. IEEE Trans. on Compulers, Sept. 1987.
[15]
P. M. Chen and D. Patterson. Maximizing performance in a striped disk array. Proc. 17th Ann. Int. Syrup. on Computer Architecture, June 1990.
[16]
S. Przybylski. The performance impact of block sizes and fetch strategies. Proc. of i7th Ann. Syrup. on Computer Arch,cture, May 1990.
[17]
K. Li and K. Petersen. Evaluation of memory system extensions. Proc. of 181h Ann. Syrup. on Computer Arch., May 1991.
Recommendations
A study of I/O system organizations
ISCA '92: Proceedings of the 19th annual international symposium on Computer architectureWith the increasing processing speeds, it has become important to design powerful and efficient I/O systems. In this paper, we look at several design options in designing an I/O system and study their impact on the performance. Specifically, we use ...
Distance-aware L2 cache organizations for scalable multiprocessor systems
Special issue: Reconfigurable embedded systems: Synthesis, design and applicationIn order to provide the scalability to the multiprocessor systems, it is important to keep the remote memory access time in bounds so that it does not impose much additional overhead as the system grows.In this paper, we suggest an LRU/distance-aware ...
Instruction Cache Replacement Policies and Organizations
Instruction cache replacement policies and organizations are analyzed both theoretically and experimentally. Theoretical analyses are based on a new model for cache referencesthe loop model. First the loop model is used to study replacement policies and ...
Comments
Please enable JavaScript to view thecomments powered by Disqus.Information & Contributors
Information
Published In
Copyright © 1992 Author.
Publisher
Association for Computing Machinery
New York, NY, United States
Publication History
Published: 01 April 1992
Published in SIGARCH Volume 20, Issue 2
Check for updates
Qualifiers
- Article
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- View Citations5Total Citations
- 446Total Downloads
- Downloads (Last 12 months)74
- Downloads (Last 6 weeks)12
Reflects downloads up to 06 Jan 2025
Other Metrics
Citations
View Options
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in