Heterogeneity-Aware Resource Provisioning Using Genetic Algorithm
Data centers have recently gained significant popularity as a cost-effective highly scalable, flexible platform to
meet emerging computing requirements and for hosting large-scale service. Data centers incurs tremendous amounts of
energy in terms of power distribution and cooling. It is challenging to effectively utilize energy in data centers. An effective
approach for saving energy in data center is to dynamically adjust data center capacity by turning off the unused machines
while matching resource demands. However, this dynamic capacity provisioning approach requires careful understanding of
heterogeneous environment and workloads. Existing heterogeneity-aware scheduling schemes rely on either trial runs or offline
profiled information to schedule the applications, which incur significant performance degradation and are impractical to
implement. To that end, we suggest a scheme that uses efficient K- means clustering algorithm to divide workload with
similar characteristics into distinct classes and genetic algorithm for scheduler.
Index terms- Cloud Computing, Resource management, Workload Characterization.