Cluster job scheduling algorithm based on multiple dimension QoS

Shan Zhao, Yunlan Wang, Xingshe Zhou

Research output: Contribution to journalArticlepeer-review

Abstract

Scheduling technology is the key of cluster job management system. At present, general scheduling used in most of cluster job management systems seldom takes the different resource requirements into consideration, which can not meet the demand of high throughout, high usability and real-time and cause the idle and overload of resources. Based on this deficiency, this paper introduces the concept of QoS, proposes the model of multiple dimension QoS and introduces a job scheduling algorithm based on MQoS(JSBMQ)according to the different application requirements and real-time resource information. This algorithm improves the system utilization and average responding time, makes the cluster provide incessant effective computing services.

Original languageEnglish
Pages (from-to)70-72
Number of pages3
JournalJisuanji Gongcheng/Computer Engineering
Volume31
Issue number22
StatePublished - 20 Nov 2005

Keywords

  • Cluster
  • Job scheduling
  • QoS

Fingerprint

Dive into the research topics of 'Cluster job scheduling algorithm based on multiple dimension QoS'. Together they form a unique fingerprint.

Cite this