Taking garbage collection overheads off the critical path in SSDs

Myoungsoo Jung, Ramya Prabhakar, Mahmut Kandemir

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

12 Citations (Scopus)

Abstract

Solid state disks (SSDs) have the potential to revolutionize the storage system landscape, mostly due to their good random access performance, compared to hard disks. However, garbage collection (GC) in SSD introduces significant latencies and large performance variations, which renders widespread adoption of SSDs difficult. To address this issue, we present a novel garbage collection strategy, consisting of two components, called Advanced Garbage Collection (AGC) and Delayed Garbage Collection (DGC), that operate collectively to migrate GC operations from busy periods to idle periods. More specifically, AGC is employed to defer GC operations to idle periods in advance, based on the type of the idle periods and on-demand GC needs, whereas DGC complements AGC by handling the collections that could not be handled by AGC. Our comprehensive experimental analysis reveals that the proposed strategies provide stable SSD performance by significantly reducing GC overheads. Compared to the state-of-the-art GC strategies, P-FTL, L-FTL and H-FTL, our AGC+DGC scheme reduces GC overheads, on average, by about 66.7%, 96.7% and 98.2%, respectively.

Original languageEnglish (US)
Title of host publicationMiddleware 2012 - ACM/IFIP/USENIX 13th International Middleware Conference, Proceedings
Pages164-186
Number of pages23
StatePublished - Nov 28 2012
Event13th ACM/IFIP/USENIX International Middleware Conference, Middleware 2012 - Montreal, QC, Canada
Duration: Dec 3 2012Dec 7 2012

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7662 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other13th ACM/IFIP/USENIX International Middleware Conference, Middleware 2012
CountryCanada
CityMontreal, QC
Period12/3/1212/7/12

Fingerprint

Garbage Collection
Critical Path
Hard disk storage
Busy Period
Random Access
Storage System
Experimental Analysis

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Jung, M., Prabhakar, R., & Kandemir, M. (2012). Taking garbage collection overheads off the critical path in SSDs. In Middleware 2012 - ACM/IFIP/USENIX 13th International Middleware Conference, Proceedings (pp. 164-186). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7662 LNCS).
Jung, Myoungsoo ; Prabhakar, Ramya ; Kandemir, Mahmut. / Taking garbage collection overheads off the critical path in SSDs. Middleware 2012 - ACM/IFIP/USENIX 13th International Middleware Conference, Proceedings. 2012. pp. 164-186 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{32029056d5e0496189f3d631ffec3ca3,
title = "Taking garbage collection overheads off the critical path in SSDs",
abstract = "Solid state disks (SSDs) have the potential to revolutionize the storage system landscape, mostly due to their good random access performance, compared to hard disks. However, garbage collection (GC) in SSD introduces significant latencies and large performance variations, which renders widespread adoption of SSDs difficult. To address this issue, we present a novel garbage collection strategy, consisting of two components, called Advanced Garbage Collection (AGC) and Delayed Garbage Collection (DGC), that operate collectively to migrate GC operations from busy periods to idle periods. More specifically, AGC is employed to defer GC operations to idle periods in advance, based on the type of the idle periods and on-demand GC needs, whereas DGC complements AGC by handling the collections that could not be handled by AGC. Our comprehensive experimental analysis reveals that the proposed strategies provide stable SSD performance by significantly reducing GC overheads. Compared to the state-of-the-art GC strategies, P-FTL, L-FTL and H-FTL, our AGC+DGC scheme reduces GC overheads, on average, by about 66.7{\%}, 96.7{\%} and 98.2{\%}, respectively.",
author = "Myoungsoo Jung and Ramya Prabhakar and Mahmut Kandemir",
year = "2012",
month = "11",
day = "28",
language = "English (US)",
isbn = "9783642351693",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "164--186",
booktitle = "Middleware 2012 - ACM/IFIP/USENIX 13th International Middleware Conference, Proceedings",

}

Jung, M, Prabhakar, R & Kandemir, M 2012, Taking garbage collection overheads off the critical path in SSDs. in Middleware 2012 - ACM/IFIP/USENIX 13th International Middleware Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7662 LNCS, pp. 164-186, 13th ACM/IFIP/USENIX International Middleware Conference, Middleware 2012, Montreal, QC, Canada, 12/3/12.

Taking garbage collection overheads off the critical path in SSDs. / Jung, Myoungsoo; Prabhakar, Ramya; Kandemir, Mahmut.

Middleware 2012 - ACM/IFIP/USENIX 13th International Middleware Conference, Proceedings. 2012. p. 164-186 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7662 LNCS).

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

TY - GEN

T1 - Taking garbage collection overheads off the critical path in SSDs

AU - Jung, Myoungsoo

AU - Prabhakar, Ramya

AU - Kandemir, Mahmut

PY - 2012/11/28

Y1 - 2012/11/28

N2 - Solid state disks (SSDs) have the potential to revolutionize the storage system landscape, mostly due to their good random access performance, compared to hard disks. However, garbage collection (GC) in SSD introduces significant latencies and large performance variations, which renders widespread adoption of SSDs difficult. To address this issue, we present a novel garbage collection strategy, consisting of two components, called Advanced Garbage Collection (AGC) and Delayed Garbage Collection (DGC), that operate collectively to migrate GC operations from busy periods to idle periods. More specifically, AGC is employed to defer GC operations to idle periods in advance, based on the type of the idle periods and on-demand GC needs, whereas DGC complements AGC by handling the collections that could not be handled by AGC. Our comprehensive experimental analysis reveals that the proposed strategies provide stable SSD performance by significantly reducing GC overheads. Compared to the state-of-the-art GC strategies, P-FTL, L-FTL and H-FTL, our AGC+DGC scheme reduces GC overheads, on average, by about 66.7%, 96.7% and 98.2%, respectively.

AB - Solid state disks (SSDs) have the potential to revolutionize the storage system landscape, mostly due to their good random access performance, compared to hard disks. However, garbage collection (GC) in SSD introduces significant latencies and large performance variations, which renders widespread adoption of SSDs difficult. To address this issue, we present a novel garbage collection strategy, consisting of two components, called Advanced Garbage Collection (AGC) and Delayed Garbage Collection (DGC), that operate collectively to migrate GC operations from busy periods to idle periods. More specifically, AGC is employed to defer GC operations to idle periods in advance, based on the type of the idle periods and on-demand GC needs, whereas DGC complements AGC by handling the collections that could not be handled by AGC. Our comprehensive experimental analysis reveals that the proposed strategies provide stable SSD performance by significantly reducing GC overheads. Compared to the state-of-the-art GC strategies, P-FTL, L-FTL and H-FTL, our AGC+DGC scheme reduces GC overheads, on average, by about 66.7%, 96.7% and 98.2%, respectively.

UR - http://www.scopus.com/inward/record.url?scp=84869834570&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84869834570&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9783642351693

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 164

EP - 186

BT - Middleware 2012 - ACM/IFIP/USENIX 13th International Middleware Conference, Proceedings

ER -

Jung M, Prabhakar R, Kandemir M. Taking garbage collection overheads off the critical path in SSDs. In Middleware 2012 - ACM/IFIP/USENIX 13th International Middleware Conference, Proceedings. 2012. p. 164-186. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).