Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Cloud management on the assumption of failure of resource demand prediction

: Uesugi, T.

Institute for Systems and Technologies of Information, Control and Communication -INSTICC-, Setubal:
2nd International Conference on Cloud Computing and Services Science, CLOSER 2012. Proceedings : Porto, Portugal, 18 - 21 April, 2012
Setúbal: SciTePress, 2012
ISBN: 978-989-8565-05-1
International Conference on Cloud Computing and Services Science (CLOSER) <2, 2012, Porto>
Fraunhofer FOKUS ()
predictive analytics; cloud computing; datacenter management

One of the important issues in cloud computing is an advanced management of large scale server clusters enabling efficient energy use and SLA compliance. That includes smart placement of virtual machines to appropriate hosts and thereby, efficient allocation of physical resources to virtual machines. One of the promising approaches is to optimize the placement based on predicting future requested physical resources for each virtual machine. However, often predictions cannot always be accurate and might cause increasing rates of SLA violation. In this paper we present an adaptive algorithm for predictive re- source allocation and optimized VM placement that offers a solution to this problem.