Improving QoS for Non-trivial Applications in Grid Computing

dc.authoridDAKKAK, OMAR/0000-0001-9767-5685
dc.authoridFazea, Yousef/0000-0003-3544-2434
dc.contributor.authorDakkak, Omar
dc.contributor.authorNor, Shahrudin Awang
dc.contributor.authorArif, Suki
dc.contributor.authorFazea, Yousef
dc.date.accessioned2024-09-29T15:50:51Z
dc.date.available2024-09-29T15:50:51Z
dc.date.issued2020
dc.departmentKarabük Üniversitesien_US
dc.description4th International Conference of Reliable Information and Communication Technology (IRICT) -- SEP 22-23, 2019 -- Johor, MALAYSIAen_US
dc.description.abstractClassical scheduling mechanisms don't satisfy the requirements for the end user, especially if the number of the jobs has increased massively in grid computing environment. To meet the expectations for non-trivial applications, the efficiency of the system has to be improved and the resources have to be ultimately utilized. Thus, backfilling technique becomes highly required due to its efficiency in exploiting the resources by filling the gaps that was created in the scheduler by short jobs. There are two well-known mechanisms, which are Extensible Argonne Scheduling System (EASY) and Conservative Backfilling (CONS). EASY is very aggressive and well uses the resources, however it causes a delay for the jobs ahead in the queue, while CONS solve this issue at the expense of system efficiency. In addition, and to further improve the scheduling quality, schedule-based approach has to be implemented. This approach provides information for the incoming job parameters and the resources capabilities; thus, the mechanism schedules the jobs in advance. This approach has shown a significant improvement compared with queue-based approach. In this paper, a new mechanism is proposed, namely Swift Gap. This mechanism implements schedule-based approach and applies multi-level scheduling method. In the first stage, the mechanism finds the right place for the newly arrival job, while in the second stage it manipulates the jobs' positions for further optimization. Moreover, this paper introduces the completion time scheme. This scheme minimizes both start time and processing time. The evaluation has shown the significant impact of Swift Gap alongside the completion time rule compared to CONS and EASY.en_US
dc.description.sponsorshipUniv Teknologi Malaysia, Informat Engn Behav Informat Res Grp, Yemeni Scientists Res Grp,Taibah Univ, Coll Comp Sci & Engn, Data Sci Res Grp,Sch Sci & Technol,Charles Darwin Univ, Coll Engn, IT & Environm,Assoc Informat Syst Malaysia Chapteren_US
dc.identifier.doi10.1007/978-3-030-33582-3_52
dc.identifier.endpage568en_US
dc.identifier.isbn978-3-030-33582-3
dc.identifier.isbn978-3-030-33581-6
dc.identifier.issn2194-5357
dc.identifier.issn2194-5365
dc.identifier.scopus2-s2.0-85077778029en_US
dc.identifier.scopusqualityN/Aen_US
dc.identifier.startpage557en_US
dc.identifier.urihttps://doi.org/10.1007/978-3-030-33582-3_52
dc.identifier.urihttps://hdl.handle.net/20.500.14619/3759
dc.identifier.volume1073en_US
dc.identifier.wosWOS:000583758100052en_US
dc.identifier.wosqualityN/Aen_US
dc.indekslendigikaynakWeb of Scienceen_US
dc.indekslendigikaynakScopusen_US
dc.language.isoenen_US
dc.publisherSpringer International Publishing Agen_US
dc.relation.ispartofEmerging Trends in Intelligent Computing and Informatics: Data Science, Intelligent Information Systems and Smart Computingen_US
dc.relation.publicationcategoryKonferans Öğesi - Uluslararası - Kurum Öğretim Elemanıen_US
dc.rightsinfo:eu-repo/semantics/closedAccessen_US
dc.subjectSchedulingen_US
dc.subjectGriden_US
dc.subjectSwift Gapen_US
dc.subjectCompletion time schemeen_US
dc.titleImproving QoS for Non-trivial Applications in Grid Computingen_US
dc.typeConference Objecten_US

Dosyalar