Predicting Parameter Sweep Jobs: From Simulation to Grid Implementation

Hellinckx, Peter; Verboven, Sam; Arickx, F.; Broeckhove, J.

doi:10.1109/cisis.2009.86

Cited by 5 publications

(1 citation statement)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The right prediction of the behavior of executing units in a computer system can be used to schedule resources more efficiently [5,6]; prediction allows the scheduler to allocate resources based on the estimated required resources. For example, if the scheduler can predict that the jobs running on a remote machine in a distributed system are about to be completed, it can rightly dispatch the next ready job to that node.…”

Section: Unpredictable Behaviormentioning

confidence: 99%

Advantages, Challenges and Optimizations of Virtual Machine Scheduling in Cloud Computing Environments

Salimi¹,

Sharifi²

2012

IJCTE

View full text Add to dashboard Cite

Abstract-Cloud computing enjoys the many attractive attributes of virtualization technology, such as consolidation, isolation, migration and suspend/resume support. In this model of computing, some desirable features such as scalability are provided by means of a new type of building blocks called virtual machines (VMs). As with any other construction block, VMs have their own scheduling challenges and advantages. This paper presents the major differences between scheduling VMs and other schedulable blocks such as processes and explains why traditional scheduling techniques used in operating systems to schedule processes or threads are not suited to VMs. New techniques to help co-scheduling virtual resources in a concurrent environment are proposed and simulated on an extension of CloudSim simulator. Simulation results of virtual processor co-scheduling show comparatively higher system performance than the methods that do not use co-scheduling.

show abstract

Section: Unpredictable Behaviormentioning

confidence: 99%