A Proposed Load Balancing Technique for Cloud Computing
Mohammed Mostafa Elmasry, Hesham
MetadataShow full item record
There are multiple challenges that influence and obstacle the rapid adoption of cloud computing infrastructure, one of these critical challenges is the low CPU utilization ratio in cloud data centers, which does not only cause a performance degradation, but also reduces the profit margin of cloud providers by raising the cost of the service provided. and leads to increasing the power consumption which causes a global warming and endangers the human life on Earth. Therefore, we need a performancepower solution to minimize the negative impact on cloud computing so as to achieve the Green Cloud Computing (GCC). To do this, the study Proposed Cloud Performance Load Balancing (CPLB) Model which can be of great help as it adopts a technique that intends to maintaining the service level agreement (SLA) and the Quality Of Service (QOS) between the cloud service providers and cloud customers and make energy conservation in the Cloud Data Center. The previous studies describing that the CPU utilization is important factor for overall system performance and cost. Greater CPU utilization produces higher response times for load dependent resources, and how the energy usage of servers in the U.S. doubled between 2000 and 2006 and how it reached to 61 billion kilowatthours (kWh) in 2006.With the fact that is only 30% of servers in data centers are fully utilized while the other 70% are still in idle state. Moreover, idle servers consume between 60% and 66% of the peak load power consumption. The study explores most of the previous load balancing techniques which are used to enhancing the cloud performance and reducing the power consumption in a cloud data center with the illustration and analysis of these techniques' shortcoming. It has been stated that all of the previous load balancing techniques focus on enhancing the response time, availability and QOS and ignore the negative impact of increasing power consumption on the surrounding environment. In order to achieve the objectives of this research and due to the nature of the study, the study used a combination of quantitate and qualitative methods, we opt to employ cloud based E-learning model, where the new CPLB model is being constructed and evaluated, we find it useful to employ the modeling and experimental methodology, because such methodology enables us to describe the proposed system model and further validate it through experiment processes, where it becomes possible to compare and analysis results between the current and proposed model.