期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2012
卷号：1
期号：5
页码：367-370
出版社：Shri Pannalal Research Institute of Technolgy
摘要：The main benefit of IaaS (Infrastructure-as-a-Service clouds) is facilitate the users to retrieve their resources on demand. But, to afford on-demand access, cloud centers must either drastically abundance their infrastructure (and pay a high price for operating resources with low utilization) or reject a large proportion of user requests (in which case the access is no longer on-demand). Several models, applications and workflows are implemented for recovering the resources where interruptions in service are predictable. For instance, many researchers and scientists utilize HTC (High Throughput Computing) enabled resources, such as Condor, where jobs are dispatched to available resources and terminated when the resource is no longer available. AS a solution we propose a special infrastructure for the cloud centers that overlaps on-demand allocation of resources with opportunistic provisioning of cycles from inactive cloud nodes to other processes by installing backfill Virtual Machines (VMs). For manifestation and tentative evaluation, we extend the Nimbus cloud computing toolkit to deploy backfill VMs on idle cloud nodes for dispensation an HTC workload. We exhibit that a mutual infrastructure amid IaaS cloud providers and an HTC job supervision system can be highly beneficial to both the IaaS cloud provider and HTC users by increasing the consumption of the cloud infrastructure and contributing cycles that would otherwise be idle to processing HTC jobs.
