DC Field | Value | Language |
---|---|---|
dc.contributor.author | Koh, Sungjoon | ko |
dc.contributor.author | Zhang, Jie | ko |
dc.contributor.author | Kwon, Miryeong | ko |
dc.contributor.author | Yoon, Jungyeon | ko |
dc.contributor.author | Donofrio, David | ko |
dc.contributor.author | Kim, Nam Sung | ko |
dc.contributor.author | Jung, Myoungsoo | ko |
dc.date.accessioned | 2019-12-13T12:30:58Z | - |
dc.date.available | 2019-12-13T12:30:58Z | - |
dc.date.created | 2019-11-28 | - |
dc.date.issued | 2017-10-01 | - |
dc.identifier.citation | 2017 IEEE International Symposium on Workload Characterization, IISWC 2017, pp.76 - 86 | - |
dc.identifier.uri | http://hdl.handle.net/10203/269606 | - |
dc.description.abstract | Large-scale systems with arrays of solid state disks (SSDs) have become increasingly common in many computing segments. To make such systems resilient, we can adopt erasure coding such as Reed-Solomon (RS) code as an alternative to replication because erasure coding can offer a significantly lower storage cost than replication. To understand the impact of using erasure coding on system performance and other system aspects such as CPU utilization and network traffic, we build a storage cluster consisting of approximately one hundred processor cores with more than fifty high-performance SSDs, and evaluate the cluster with a popular open-source distributed parallel file system, Ceph. Then we analyze behaviors of systems adopting erasure coding from the following five viewpoints, compared with those of systems using replication: (1) storage system I/O performance; (2) computing and software overheads; (3) I/O amplification; (4) network traffic among storage nodes; (5) the impact of physical data layout on performance of RS-coded SSD arrays. For all these analyses, we examine two representative RS configurations, which are used by Google and Facebook file systems, and compare them with triple replication that a typical parallel file system employs as a default fault tolerance mechanism. Lastly, we collect 54 block-level traces from the cluster and make them available for other researchers. | - |
dc.language | English | - |
dc.publisher | IEEE Computer Society | - |
dc.title | Understanding system characteristics of online erasure coding on scalable, distributed and large-scale SSD array systems | - |
dc.type | Conference | - |
dc.type.rims | CONF | - |
dc.citation.beginningpage | 76 | - |
dc.citation.endingpage | 86 | - |
dc.citation.publicationname | 2017 IEEE International Symposium on Workload Characterization, IISWC 2017 | - |
dc.identifier.conferencecountry | US | - |
dc.identifier.conferencelocation | Seattle, Washington | - |
dc.identifier.doi | 10.1109/IISWC.2017.8167758 | - |
dc.contributor.localauthor | Jung, Myoungsoo | - |
dc.contributor.nonIdAuthor | Koh, Sungjoon | - |
dc.contributor.nonIdAuthor | Zhang, Jie | - |
dc.contributor.nonIdAuthor | Kwon, Miryeong | - |
dc.contributor.nonIdAuthor | Yoon, Jungyeon | - |
dc.contributor.nonIdAuthor | Donofrio, David | - |
dc.contributor.nonIdAuthor | Kim, Nam Sung | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.