StorkCloud: Data transfer scheduling and optimization as a service

Tevfik Kosar, Engin Arslan, Brandon Ross, Bing Zhang

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

Abstract

Wide-area transfer of large data sets is still a big challenge despite the deployment of high-bandwidth networks with speeds reaching 100 Gbps. Most users fail to obtain even a fraction of theoretical speeds promised by these networks. Effective usage of the available network capacity has become increasingly important for wide-area data movement. We have developed a "data transfer scheduling and optimization system as a Cloud-hosted service", StorkCloud, which will mitigate the large-scale end-to-end data movement bottleneck by efficiently utilizing underlying networks and effectively scheduling and optimizing data transfers. In this paper, we present the initial design and prototype implementation of StorkCloud, and show its effectiveness in large dataset transfers across geographically distant storage sites, data centers, and collaborating institutions.

Original languageEnglish (US)
Title of host publicationScienceCloud 2013 - Proceedings of the 4th ACM Workshop on Scientific Cloud Computing
PublisherAssociation for Computing Machinery
Pages29-36
Number of pages8
ISBN (Print)9781450319799
DOIs
StatePublished - Jan 1 2013
Externally publishedYes
Event4th ACM Workshop on Scientific Cloud Computing, ScienceCloud 2013 - New York, NY, United States
Duration: Jun 17 2013Jun 17 2013

Publication series

NameScienceCloud 2013 - Proceedings of the 4th ACM Workshop on Scientific Cloud Computing

Conference

Conference4th ACM Workshop on Scientific Cloud Computing, ScienceCloud 2013
CountryUnited States
CityNew York, NY
Period6/17/136/17/13

Keywords

  • big data
  • cloud computing
  • data scheduling.
  • end-to-end throughput optimization
  • software as a service (SAAS)

ASJC Scopus subject areas

  • Software

Fingerprint Dive into the research topics of 'StorkCloud: Data transfer scheduling and optimization as a service'. Together they form a unique fingerprint.

Cite this