File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)

Conference Paper: Delay performance of direct reads in distributed storage systems with coding

TitleDelay performance of direct reads in distributed storage systems with coding
Authors
KeywordsBandwidth cost
Degraded reads
Delay
Direct reads
Hot data
Load balancing
MDS codes
Issue Date2015
PublisherIEEE Computer Society. The Journal's web site is located at http://ieeexplore.ieee.org/xpl/conhome.jsp?punumber=1002461
Citation
The IEEE 17th International Conference on High Performance Computing and Communications (HPCC), IEEE 7th International Symposium on Cyberspace Safety and Security (CSS), IEEE 12th International Conference on Embedded Software and Systems (ICESS), New York, NY., 24-26 August 2015. In Conference Proceedings, 2015, p. 184-189 How to Cite?
AbstractDelay performance is crucial in distributed storage systems and it can greatly impact user experience. Both direct and k-access reads are common in storage systems. However, much of previous research only considers k-access reads and many schemes, such as Redundant Scheme, are only shown to reduce delay for k-access reads. We have no idea whether those existing schemes can also work for direct reads. The study regarding the characteristics of the delay performance ofdirect reads, and the appropriate schemes for direct reads to reduce delay is still lacking. In this paper, we study the delay performance of direct reads and its correlation with degraded reads. We illustrate the relationship between degraded reads and bandwidth cost and answer important questions like when degraded reads can help reduce delay. Then we propose a scheme DRALB to reduce delay for direct reads. DRALB can be easily added to existing schemes and can greatly reduce the delay of hot data. We also conduct trace-driven simulations to verify that DRALB significantly outperforms existing schemes, in terms of delay performance of direct reads. © 2015 IEEE.
Persistent Identifierhttp://hdl.handle.net/10722/232306
ISBN

 

DC FieldValueLanguage
dc.contributor.authorShuai, Q-
dc.contributor.authorLi, VOK-
dc.date.accessioned2016-09-20T05:29:06Z-
dc.date.available2016-09-20T05:29:06Z-
dc.date.issued2015-
dc.identifier.citationThe IEEE 17th International Conference on High Performance Computing and Communications (HPCC), IEEE 7th International Symposium on Cyberspace Safety and Security (CSS), IEEE 12th International Conference on Embedded Software and Systems (ICESS), New York, NY., 24-26 August 2015. In Conference Proceedings, 2015, p. 184-189-
dc.identifier.isbn978-147998936-2-
dc.identifier.urihttp://hdl.handle.net/10722/232306-
dc.description.abstractDelay performance is crucial in distributed storage systems and it can greatly impact user experience. Both direct and k-access reads are common in storage systems. However, much of previous research only considers k-access reads and many schemes, such as Redundant Scheme, are only shown to reduce delay for k-access reads. We have no idea whether those existing schemes can also work for direct reads. The study regarding the characteristics of the delay performance ofdirect reads, and the appropriate schemes for direct reads to reduce delay is still lacking. In this paper, we study the delay performance of direct reads and its correlation with degraded reads. We illustrate the relationship between degraded reads and bandwidth cost and answer important questions like when degraded reads can help reduce delay. Then we propose a scheme DRALB to reduce delay for direct reads. DRALB can be easily added to existing schemes and can greatly reduce the delay of hot data. We also conduct trace-driven simulations to verify that DRALB significantly outperforms existing schemes, in terms of delay performance of direct reads. © 2015 IEEE.-
dc.languageeng-
dc.publisherIEEE Computer Society. The Journal's web site is located at http://ieeexplore.ieee.org/xpl/conhome.jsp?punumber=1002461-
dc.relation.ispartofIEEE International Conference on High Performance Computing and Communications Proceedings-
dc.subjectBandwidth cost-
dc.subjectDegraded reads-
dc.subjectDelay-
dc.subjectDirect reads-
dc.subjectHot data-
dc.subjectLoad balancing-
dc.subjectMDS codes-
dc.titleDelay performance of direct reads in distributed storage systems with coding-
dc.typeConference_Paper-
dc.identifier.emailLi, VOK: vli@eee.hku.hk-
dc.identifier.authorityLi, VOK=rp00150-
dc.description.natureLink_to_subscribed_fulltext-
dc.identifier.doi10.1109/HPCC-CSS-ICESS.2015.200-
dc.identifier.scopuseid_2-s2.0-84961695490-
dc.identifier.hkuros265295-
dc.identifier.spage184-
dc.identifier.epage189-
dc.publisher.placeUnited States-
dc.customcontrol.immutablesml 161003-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats