A virtual machine consolidation framework for MapReduce enabled computing clouds

Authors:
Zhe Huang;Danny H. K. Tsang;James She
Affiliations:
The Hong Kong University of Science and Technology;The Hong Kong University of Science and Technology;The Hong Kong University of Science and Technology
Venue:
Proceedings of the 24th International Teletraffic Congress
Year:
2012

Citing 6
Cited 0

MapReduce: simplified data processing on large clusters

Communications of the ACM - 50th anniversary issue: 1958 - 2008
CloudBLAST: Combining MapReduce and Virtualization on Distributed Resources for Bioinformatics Applications

ESCIENCE '08 Proceedings of the 2008 Fourth IEEE International Conference on eScience
Hadoop: The Definitive Guide

Hadoop: The Definitive Guide
Efficient resource provisioning in compute clouds via VM multiplexing

Proceedings of the 7th international conference on Autonomic computing
Improving MapReduce performance in heterogeneous environments

OSDI'08 Proceedings of the 8th USENIX conference on Operating systems design and implementation
MapReduce in the Clouds for Science

CLOUDCOM '10 Proceedings of the 2010 IEEE Second International Conference on Cloud Computing Technology and Science

Quantified Score

Hi-index	0.00

Visualization

Abstract

In nowadays computing clouds, it is of the cloud providers' economic interests to correctly consolidate the workload of the virtual machines (VMs) into the suitable physical servers in the cloud data center in order to minimize the total maintenance cost. However, during the consolidation process, sufficient protection should be provided to the service level agreement (SLA) of the VMs. In this paper, the VM consolidation problem for MapReduce enabled computing clouds has been investigated. In the MapReduce enabled computing clouds, MapReduce jobs are carried out by homogeneous MapReduce VM instances that have identical hardware resource. Two resource allocation schemes with corresponding SLA constraints for the MapReduce VMs and the non-MapReduce VMs are proposed. Based on these schemes, the VM consolidation problem is modeled as an integer nonlinear optimization problem and an efficient algorithm has been proposed to locate its solutions. The results show that better VM consolidation performance can be achieved by colocating MapReduce instances together with non-MapReduce instances in the same set of physical servers.