ABSTRACT

Approximately 55 percent of the world’s 7.3 billion people access the Internet, creating a significant demand in information technology (IT) services for both organizations and consumers. The number of data centers continue to increase to meet this demand, consuming 2 percent of the energy produced worldwide. Many organizations moving towards cloud computing due to its ability to meet users’ needs on demand. Greening cloud computing technology has become an important aspect of an organization’s design and use of cloud computing. One aspect of greening the cloud is through efficiently using cloud-based resources. In this study, we focus on the resource allocation applications of cloud computing technologies to green an organizations cloud. We design a pricing and allocation mechanism for a private cloud computing service that allows the firm to effectively load balance their cloud computing resources. Our optimal pricing mechanism is a dynamic pricing model, which maximizes the net value of users in a private cloud computing service. Additionally, we create a job allocation algorithm based on our dynamic pricing model to load balance the cloud. We show that balancing the job allocation such that the number of jobs at individual resource servers as close to equal as possible is optimal. Furthermore, simulations were run on the allocation mechanism to examine the effects on the cloud resources while gaining insight to effectively distribution resources to public clouds. Our model can help organizations to efficiently distribute their cloud-based resources, which allows for a greener cloud computing system.

摘要

全球 73 亿人中约有 55% 访问互联网，这为组织和消费者创造了对信息技术 (IT) 服务的巨大需求。数据中心的数量不断增加以满足这一需求，消耗了全球 2% 的能源。许多组织转向云计算是因为它能够按需满足用户的需求。绿化云计算技术已成为组织设计和使用云计算的重要方面。绿化云的一个方面是通过有效地使用基于云的资源。在本研究中，我们专注于云计算技术的资源分配应用，以实现组织云的绿色化。我们为私有云计算服务设计了一种定价和分配机制，使公司能够有效地平衡其云计算资源的负载。我们的最优定价机制是动态定价模型，在私有云计算服务中最大化用户的净价值。此外，我们基于我们的动态定价模型创建了一个作业分配算法来平衡云。我们表明，平衡作业分配以使各个资源服务器上的作业数量尽可能接近相等是最佳的。此外，对分配机制进行了模拟，以检查对云资源的影响，同时深入了解如何有效地将资源分配到公共​​云。我们的模型可以帮助组织有效地分配其基于云的资源，