On urgency of I/O operations

Mahmut Kandemir, Taylan Yemliha, Ramya Prabhakar, Myoungsoo Jung

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Many high-performance parallel file systems and storage hierarchies employ multilayer storage caches in an attempt to reduce data access latencies. In current storage cache hierarchies, all data requests are treated uniformly and hit/miss characteristics are dictated only by the degree of reuse exhibited by data blocks. In reality however, different I/O operations may have different urgencies (criticalities), and in particular, some I/O operations can be delayed without having a major impact on overall application performance. Motivated by this observation, we define the concept of I/O operation urgency (criticality) and study the critical latencies of I/O operations for a set of seven high-performance applications that manipulate disk-resident data sets. We propose and experimentally evaluate three profile-based strategies for exploiting urgent I/O operations in managing storage caches. The results collected with these schemes on both two-tier and three-tier systems indicate that significant performance improvements are possible if one could exploit urgencies of different I/O operations in managing storage caches.

Original languageEnglish (US)
Title of host publicationProceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012
Pages188-195
Number of pages8
DOIs
StatePublished - Jul 16 2012
Event12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012 - Ottawa, ON, Canada
Duration: May 13 2012May 16 2012

Publication series

NameProceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012

Other

Other12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012
CountryCanada
CityOttawa, ON
Period5/13/125/16/12

Fingerprint

Cache
Multilayers
Criticality
Latency
High Performance
Parallel File System
Hits
Reuse
Multilayer
Evaluate
Hierarchy

All Science Journal Classification (ASJC) codes

  • Computational Theory and Mathematics
  • Applied Mathematics

Cite this

Kandemir, M., Yemliha, T., Prabhakar, R., & Jung, M. (2012). On urgency of I/O operations. In Proceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012 (pp. 188-195). [6217421] (Proceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012). https://doi.org/10.1109/CCGrid.2012.40
Kandemir, Mahmut ; Yemliha, Taylan ; Prabhakar, Ramya ; Jung, Myoungsoo. / On urgency of I/O operations. Proceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012. 2012. pp. 188-195 (Proceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012).
@inproceedings{b5ccae1682a34a9697efa30369d7f2d9,
title = "On urgency of I/O operations",
abstract = "Many high-performance parallel file systems and storage hierarchies employ multilayer storage caches in an attempt to reduce data access latencies. In current storage cache hierarchies, all data requests are treated uniformly and hit/miss characteristics are dictated only by the degree of reuse exhibited by data blocks. In reality however, different I/O operations may have different urgencies (criticalities), and in particular, some I/O operations can be delayed without having a major impact on overall application performance. Motivated by this observation, we define the concept of I/O operation urgency (criticality) and study the critical latencies of I/O operations for a set of seven high-performance applications that manipulate disk-resident data sets. We propose and experimentally evaluate three profile-based strategies for exploiting urgent I/O operations in managing storage caches. The results collected with these schemes on both two-tier and three-tier systems indicate that significant performance improvements are possible if one could exploit urgencies of different I/O operations in managing storage caches.",
author = "Mahmut Kandemir and Taylan Yemliha and Ramya Prabhakar and Myoungsoo Jung",
year = "2012",
month = "7",
day = "16",
doi = "10.1109/CCGrid.2012.40",
language = "English (US)",
isbn = "9780769546919",
series = "Proceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012",
pages = "188--195",
booktitle = "Proceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012",

}

Kandemir, M, Yemliha, T, Prabhakar, R & Jung, M 2012, On urgency of I/O operations. in Proceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012., 6217421, Proceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012, pp. 188-195, 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012, Ottawa, ON, Canada, 5/13/12. https://doi.org/10.1109/CCGrid.2012.40

On urgency of I/O operations. / Kandemir, Mahmut; Yemliha, Taylan; Prabhakar, Ramya; Jung, Myoungsoo.

Proceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012. 2012. p. 188-195 6217421 (Proceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - On urgency of I/O operations

AU - Kandemir, Mahmut

AU - Yemliha, Taylan

AU - Prabhakar, Ramya

AU - Jung, Myoungsoo

PY - 2012/7/16

Y1 - 2012/7/16

N2 - Many high-performance parallel file systems and storage hierarchies employ multilayer storage caches in an attempt to reduce data access latencies. In current storage cache hierarchies, all data requests are treated uniformly and hit/miss characteristics are dictated only by the degree of reuse exhibited by data blocks. In reality however, different I/O operations may have different urgencies (criticalities), and in particular, some I/O operations can be delayed without having a major impact on overall application performance. Motivated by this observation, we define the concept of I/O operation urgency (criticality) and study the critical latencies of I/O operations for a set of seven high-performance applications that manipulate disk-resident data sets. We propose and experimentally evaluate three profile-based strategies for exploiting urgent I/O operations in managing storage caches. The results collected with these schemes on both two-tier and three-tier systems indicate that significant performance improvements are possible if one could exploit urgencies of different I/O operations in managing storage caches.

AB - Many high-performance parallel file systems and storage hierarchies employ multilayer storage caches in an attempt to reduce data access latencies. In current storage cache hierarchies, all data requests are treated uniformly and hit/miss characteristics are dictated only by the degree of reuse exhibited by data blocks. In reality however, different I/O operations may have different urgencies (criticalities), and in particular, some I/O operations can be delayed without having a major impact on overall application performance. Motivated by this observation, we define the concept of I/O operation urgency (criticality) and study the critical latencies of I/O operations for a set of seven high-performance applications that manipulate disk-resident data sets. We propose and experimentally evaluate three profile-based strategies for exploiting urgent I/O operations in managing storage caches. The results collected with these schemes on both two-tier and three-tier systems indicate that significant performance improvements are possible if one could exploit urgencies of different I/O operations in managing storage caches.

UR - http://www.scopus.com/inward/record.url?scp=84863698667&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84863698667&partnerID=8YFLogxK

U2 - 10.1109/CCGrid.2012.40

DO - 10.1109/CCGrid.2012.40

M3 - Conference contribution

SN - 9780769546919

T3 - Proceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012

SP - 188

EP - 195

BT - Proceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012

ER -

Kandemir M, Yemliha T, Prabhakar R, Jung M. On urgency of I/O operations. In Proceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012. 2012. p. 188-195. 6217421. (Proceedings - 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGrid 2012). https://doi.org/10.1109/CCGrid.2012.40