No abstract

The problem of balancing user requests in cloud computing is becoming more serious due to the variation of workloads. Load balancing and allocation processes still need more optimizing methodologies and models to improve performance and increase the quality of service. This article describes a solution to balance user workload efficiently by proposing a model that allows each virtual machine (VM) to maximize the serving number of requests based on its capacity. The model measures VMs' capacity as a percentage and maps groups of user requests to appropriate active virtual machines. Finding the expected patterns from a big data repository, such as log data, and using some machine learning techniques can make the prediction more efficiently. The work is implemented and evaluated using some performance metrics, and the results are compared with other research. The evaluation shows the efficiency of the proposed approach in distributing user workload and improving results.

Cloud computing provides different services through data centers that are often located in different geographical locations. The users are faced with a wide variety of services to choose from. Also, with the increasing number of serviced applications, overloading might occur on service brokers for balancing and serving the requests. Consequently, maximizing the number of entry points and considering the maximum number of factors that affect the performance for balancing the workload is very important for the quality of service. This paper proposes a model named multi-cloud service brokers (MCSB) for selecting the optimal DC using multiple entry points. The developed service broker policy shares information about the requests considering some new performance factors. This extension is added to the CloudAnalyst simulator tool which is used in this work, and the results are evaluated and compared to other existing policies from the literature.