Comma

Abstract

Virtualization technology has made possible, providing memory, disk and processing power to consumers based on pre-defined conditions to be applied on both cloud consumers and cloud hosts. A consumer can have many Virtual Machines (VMs) at a time and a cloud host may have many such consumers at the same time. Competition in Cloud Service Providers (CSPs) and to maintain service level agreements, it is necessary for a CSP to have market-oriented leasing policy which can provide higher profit to cloud host while maintaining lowest cost, shortest downtime and maximum availability of resources to prioritized consumers at peak times. This paper introduces a new leasing policy and algorithm: COMMA (Cost Oriented, Market and Migration Aware) for maintaining a balance between cost, profit and migrations of VM during peak loads. Experiments and results show that this leasing policy and algorithm are effective and can be used for making profit and maintain QoS in a cloud environment.