Disk-cache and parallelism aware I/O scheduling to improve storage system performance

Ramya Prabhakar, Mahmut Kandemir, Myoungsoo Jung

Research output: Contribution to conferencePaper

3 Citations (Scopus)

Abstract

Modern large computing systems employ sophisticated disk I/O systems that are configured to deliver high throughput, low-latency disk I/O to multiple clients accessing them. However, due to potential interferences among concurrent I/O accesses issued by multiple clients, a disk-cache and disk-level parallelism unaware I/O scheduling algorithm employed by the operating system/storage controller may have a significant impact on both system throughput and I/O latency. In this paper, we propose two fundamentally new disk I/O scheduling techniques. The first technique, called DCAP, performs I/O scheduling in a disk cache aware and parallelism aware manner. The key idea in DCAP is to process simultaneous requests to different disks from the same application/priority class together and reorder them so that they have the highest number of hits in the disk cache. We then propose an enhanced version of DCAP called DCAP-G, that aggregates requests into service groups to alleviate the problem of request starvation that may occur in DCAP in certain cases. We evaluate both DCAP and DCAP-G using a set of I/O workloads from production-based enterprise systems as well as high-performance computing domain. In addition, we also compare the performance of our algorithms to previously proposed I/O scheduling algorithms. Our evaluation shows that, averaged across all our workloads, DCAP improves the average I/O response time, taking maximum advantage of disk access locality and exploiting parallelism among concurrent accesses to multiple disks, by 14.9% over an I/O scheduler that schedules requests on a first-come-first-served (FCFS) basis and also improves by 6.5% over a previously proposed locality-optimal I/O scheduler (SPCTF). In addition to these improvements, DCAP-G improves the average I/O response time by 6.6% over DCAP, leading to an overall 20.7% and 12.0% improvement over FCFS, and SPCTF, respectively.

Original languageEnglish
Pages357-368
Number of pages12
DOIs
Publication statusPublished - 2013 Oct 7
Event27th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2013 - Boston, MA, United States
Duration: 2013 May 202013 May 24

Other

Other27th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2013
CountryUnited States
CityBoston, MA
Period13/5/2013/5/24

Fingerprint

Scheduling algorithms
Scheduling
Throughput
Controllers
Industry

All Science Journal Classification (ASJC) codes

  • Software

Cite this

Prabhakar, R., Kandemir, M., & Jung, M. (2013). Disk-cache and parallelism aware I/O scheduling to improve storage system performance. 357-368. Paper presented at 27th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2013, Boston, MA, United States. https://doi.org/10.1109/IPDPS.2013.59
Prabhakar, Ramya ; Kandemir, Mahmut ; Jung, Myoungsoo. / Disk-cache and parallelism aware I/O scheduling to improve storage system performance. Paper presented at 27th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2013, Boston, MA, United States.12 p.
@conference{315de49c4bd0409888620a1fa5125dd2,
title = "Disk-cache and parallelism aware I/O scheduling to improve storage system performance",
abstract = "Modern large computing systems employ sophisticated disk I/O systems that are configured to deliver high throughput, low-latency disk I/O to multiple clients accessing them. However, due to potential interferences among concurrent I/O accesses issued by multiple clients, a disk-cache and disk-level parallelism unaware I/O scheduling algorithm employed by the operating system/storage controller may have a significant impact on both system throughput and I/O latency. In this paper, we propose two fundamentally new disk I/O scheduling techniques. The first technique, called DCAP, performs I/O scheduling in a disk cache aware and parallelism aware manner. The key idea in DCAP is to process simultaneous requests to different disks from the same application/priority class together and reorder them so that they have the highest number of hits in the disk cache. We then propose an enhanced version of DCAP called DCAP-G, that aggregates requests into service groups to alleviate the problem of request starvation that may occur in DCAP in certain cases. We evaluate both DCAP and DCAP-G using a set of I/O workloads from production-based enterprise systems as well as high-performance computing domain. In addition, we also compare the performance of our algorithms to previously proposed I/O scheduling algorithms. Our evaluation shows that, averaged across all our workloads, DCAP improves the average I/O response time, taking maximum advantage of disk access locality and exploiting parallelism among concurrent accesses to multiple disks, by 14.9{\%} over an I/O scheduler that schedules requests on a first-come-first-served (FCFS) basis and also improves by 6.5{\%} over a previously proposed locality-optimal I/O scheduler (SPCTF). In addition to these improvements, DCAP-G improves the average I/O response time by 6.6{\%} over DCAP, leading to an overall 20.7{\%} and 12.0{\%} improvement over FCFS, and SPCTF, respectively.",
author = "Ramya Prabhakar and Mahmut Kandemir and Myoungsoo Jung",
year = "2013",
month = "10",
day = "7",
doi = "10.1109/IPDPS.2013.59",
language = "English",
pages = "357--368",
note = "27th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2013 ; Conference date: 20-05-2013 Through 24-05-2013",

}

Prabhakar, R, Kandemir, M & Jung, M 2013, 'Disk-cache and parallelism aware I/O scheduling to improve storage system performance' Paper presented at 27th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2013, Boston, MA, United States, 13/5/20 - 13/5/24, pp. 357-368. https://doi.org/10.1109/IPDPS.2013.59

Disk-cache and parallelism aware I/O scheduling to improve storage system performance. / Prabhakar, Ramya; Kandemir, Mahmut; Jung, Myoungsoo.

2013. 357-368 Paper presented at 27th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2013, Boston, MA, United States.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Disk-cache and parallelism aware I/O scheduling to improve storage system performance

AU - Prabhakar, Ramya

AU - Kandemir, Mahmut

AU - Jung, Myoungsoo

PY - 2013/10/7

Y1 - 2013/10/7

N2 - Modern large computing systems employ sophisticated disk I/O systems that are configured to deliver high throughput, low-latency disk I/O to multiple clients accessing them. However, due to potential interferences among concurrent I/O accesses issued by multiple clients, a disk-cache and disk-level parallelism unaware I/O scheduling algorithm employed by the operating system/storage controller may have a significant impact on both system throughput and I/O latency. In this paper, we propose two fundamentally new disk I/O scheduling techniques. The first technique, called DCAP, performs I/O scheduling in a disk cache aware and parallelism aware manner. The key idea in DCAP is to process simultaneous requests to different disks from the same application/priority class together and reorder them so that they have the highest number of hits in the disk cache. We then propose an enhanced version of DCAP called DCAP-G, that aggregates requests into service groups to alleviate the problem of request starvation that may occur in DCAP in certain cases. We evaluate both DCAP and DCAP-G using a set of I/O workloads from production-based enterprise systems as well as high-performance computing domain. In addition, we also compare the performance of our algorithms to previously proposed I/O scheduling algorithms. Our evaluation shows that, averaged across all our workloads, DCAP improves the average I/O response time, taking maximum advantage of disk access locality and exploiting parallelism among concurrent accesses to multiple disks, by 14.9% over an I/O scheduler that schedules requests on a first-come-first-served (FCFS) basis and also improves by 6.5% over a previously proposed locality-optimal I/O scheduler (SPCTF). In addition to these improvements, DCAP-G improves the average I/O response time by 6.6% over DCAP, leading to an overall 20.7% and 12.0% improvement over FCFS, and SPCTF, respectively.

AB - Modern large computing systems employ sophisticated disk I/O systems that are configured to deliver high throughput, low-latency disk I/O to multiple clients accessing them. However, due to potential interferences among concurrent I/O accesses issued by multiple clients, a disk-cache and disk-level parallelism unaware I/O scheduling algorithm employed by the operating system/storage controller may have a significant impact on both system throughput and I/O latency. In this paper, we propose two fundamentally new disk I/O scheduling techniques. The first technique, called DCAP, performs I/O scheduling in a disk cache aware and parallelism aware manner. The key idea in DCAP is to process simultaneous requests to different disks from the same application/priority class together and reorder them so that they have the highest number of hits in the disk cache. We then propose an enhanced version of DCAP called DCAP-G, that aggregates requests into service groups to alleviate the problem of request starvation that may occur in DCAP in certain cases. We evaluate both DCAP and DCAP-G using a set of I/O workloads from production-based enterprise systems as well as high-performance computing domain. In addition, we also compare the performance of our algorithms to previously proposed I/O scheduling algorithms. Our evaluation shows that, averaged across all our workloads, DCAP improves the average I/O response time, taking maximum advantage of disk access locality and exploiting parallelism among concurrent accesses to multiple disks, by 14.9% over an I/O scheduler that schedules requests on a first-come-first-served (FCFS) basis and also improves by 6.5% over a previously proposed locality-optimal I/O scheduler (SPCTF). In addition to these improvements, DCAP-G improves the average I/O response time by 6.6% over DCAP, leading to an overall 20.7% and 12.0% improvement over FCFS, and SPCTF, respectively.

UR - http://www.scopus.com/inward/record.url?scp=84884844582&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84884844582&partnerID=8YFLogxK

U2 - 10.1109/IPDPS.2013.59

DO - 10.1109/IPDPS.2013.59

M3 - Paper

AN - SCOPUS:84884844582

SP - 357

EP - 368

ER -

Prabhakar R, Kandemir M, Jung M. Disk-cache and parallelism aware I/O scheduling to improve storage system performance. 2013. Paper presented at 27th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2013, Boston, MA, United States. https://doi.org/10.1109/IPDPS.2013.59