Dynamic virtual machine allocation based on adaptive genetic algorithm
Author
Rolik, Oleksandr
Telenyk, Sergii
Zharikov, Eduard
Samotyy, Volodymyr
Published in
CLOUD COMPUTING 2017: The Eighth International Conference on Cloud Computing, GRIDs, and Virtualization
Numbering
1
Pages
108-114
Release date
2017
Place of publication
Athens, Greece
Publisher
IARIA
Language
English
ISSN
2308-4294
ISBN
978-1-61208-529-6
Keywords
data center, genetic algorithm, virtual machine, resource management
Abstract
The widespread use of the virtualization paradigm in modern data centers has increased the necessity of improving the management efficiency of virtual machine allocation on physical machines (PM). Modern service providers offer a large number of virtual machine types and settings. The density of virtual machine placement per physical server also complicates the solution of this problem. Under these conditions, for solving such kind of problems, the adaptive genetic algorithm (AGA) is proposed. The proposed algorithm uses parametric and algorithmic adaptation simultaneously by selecting the values for a genetic operator’s parameters and by selecting the probabilities of applying these operators. The AGA is evaluated for the solution of virtual machine allocation problem and demonstrates efficiency compared to the classical and the controlled versions of genetic algorithm.