Towards optimization of availability and cost in selection of geo-distributed clouds datacenter

With increasing data clouds in different geographical areas, the availability of a datacenter and the cost of using the datacenter are two concerned factors of clouds users. The present research aims to present a method using K-means clustering and NSGA-II multi-objective algorithm to maximize availability and minimizes cost in selecting a datacenter. The proposed approach was applied to some real geographically distributed datacenters. Results showed that proposed approach outperforms greedy and random common algorithms.