A Linear Programming Approach for Optimizing Workload Distribution in a Cloud

Cloud computing's usage-based pricing model creates an incentive for subscribers to optimize the utilization of the rented resources. The goal of the current work is to devise a formal approach for distributing workload among a minimum number of servers. The paper models this problem as a set partitioning problem and describes two solution approaches. The first one generates a set of candidate blocks and then composes an optimal partition by solving an integer programming problem. The second approach solves the set partitioning problem with column generation technique. Both methods were implemented and evaluated. The experiment results led to a conclusion that the second approach delivers the best results.