Title :
Sharing-Aware Online Algorithms for Virtual Machine Packing in Cloud Environments
Author :
Rampersaud, Safraz ; Grosu, Daniel
Author_Institution :
Dept. of Comput. Sci., Wayne State Univ., Detroit, MI, USA
Abstract :
Cloud service providers offer services to a large number of users by employing virtualization technologies. One of the challenges faced by the cloud providers using virtualized environments is the development of efficient algorithms for assigning Virtual Machine (VM) instances to servers such that the number of hosting servers is minimized. This is also known as the problem of VM Packing. In this paper, we design a family of sharing-aware online algorithms for solving the VM Packing problem that take into account the sharing of memory among collocated VMs. We introduce a new server resource scarcity metric which establishes an order among servers that are suitable for hosting an arriving VM request and use it in the design of the algorithms. We evaluate the performance of our sharing-aware online algorithms by performing an extensive set of experiments comparing them against several existing sharing-oblivious VM packing algorithms.
Keywords :
cloud computing; virtual machines; virtualisation; cloud environments; cloud service providers; sharing-aware online algorithm; virtual machine packing; Algorithm design and analysis; Measurement; Memory management; Random access memory; Resource management; Servers; Silicon; online algorithms; sharing-aware VM packing; vector bin packing; virtual machines;
Conference_Titel :
Cloud Computing (CLOUD), 2015 IEEE 8th International Conference on
Conference_Location :
New York City, NY
Print_ISBN :
978-1-4673-7286-2
DOI :
10.1109/CLOUD.2015.100