مقاله انگلیسی رایگان در مورد تخصیص منابع مشتری براساس GA – اسپرینگر 2017

Multi-cloud computing is a technical paradigm, which provides resources of various capacities to customers. Each job request receives unique job id after registration in the multi-cloud federation. For day-to-day enterprise operation, health, military, IT industry operations, etc., cloud computing is inevitable. The varied features of multi-cloud computing like on-demand services, pay-as-you-go pricing model, dynamic scaling, virtualisation and no vendor lock-in make cloud platform appealing for industries of all capacities and research society. The cloud services are provisioned in the form of storage datacentre. However, Infrastructure as a Service (IaaS) cloud uses a scheduling policy to allocate VMs to the customer requests. Amazon Web Services (AWS) use First In First Out, scheduling using batch processing. The success of cloud computing relies on load balancing, efficient scheduling and importantly collaborating among peer cloud service providers to form a reliable federation to provide complex problem-solving techniques handling day-to-day business, scientific and engineering applications. No datacentre provides unlimited resources to accommodate dynamic scaling. Customer having high demanding applications reserves higher instances to get the timely deployment. The multi-cloud computing has its own share of challenges. In multi-cloud computing platform optimising numerous resources which are distributed in a different geographical region is challenging. Basically, a centralised management is established to achieve resource allocation followed by task scheduling. Centralised cloud broker monitors the fitness, status of VMs and the scheduling procedures. Assigning available tasks to VMs is one of the concerned problem which has taken attention from academia, business and research, assigning VMs to include two-phase processes, namely mapping and scheduling. Mapping includes allo cating the incoming tasks to processing units termed VMS, followed by arranging the order of the allocated tasks. In this paper, we propose GA-based Customer-Conscious Resource Allocation and Task Scheduling (GACCRATS). The algorithm is divided into two crucial phases, namely resource allocation and task scheduling. The experimental outcomes prove that the proposed model outperforms the existing algorithms in terms of makespan time and customer satisfaction rate in the multi-cloud platform.