Improved scheduling using meta heuristic self adaptive group formation in cloud

Abstract

Cloud computing is a modern technology, which is required for sharing the computational resources between the computers to reduce the scale of economy. The cloud computing shares the resources via public or private utility to reduce the maintenance of high level computer infrastructure. The cloud computing reduces the front end infrastructure and enables the system to work faster with flexible database storage. The cloud computing can also be considered as an alternate process to improve the production of business and organization. The cloud is usually constructed using the cloud provider through various virtual machines from several service providers and it tends to process each tasks of users. However, in order to provide flexible access, a non-submissive model is presented and hence the users achieve the services in a faster way with low effort using the required configurable resources. The cloud computing scheduler schedules the user task through several high performance cloud computing algorithms. The scheduling of user task should consider the application and multiple user sharing the cloud resources, system maintenance and resources utilization. However, maintaining these four factors is considered to be a difficult task. There are other constraints that affects the scheduling efficiency of cloud computing, that includes processor power, memory space and network bandwidth. Moreover, scheduler that schedules the task through several cloud computing resources increases the computational cost of cloud VMs. If the exchange of data is rapid, the scheduling tends to become computationally intensive between the nodes, host and clusters. There exist several cloud computing models that operates on CPU allocation and memory resources allocation to schedule various tasks by considering that all physical nodes and VMs has unlimited resources or bandwidth. Hence, the scheduling algorithm is used to manage the multidimensional task based on the availability of CPU, availability of memory and availability of VMs. Further, the resources limitation is also considered and it allocates the resources based on the resource loads and the requirement of tasks. It could be seen that while considering the bandwidth requirement as a major constraints, there exist a limitation of resource requirement during task allocation. The task scheduling is further limited by the CPU requirement, bandwidth utilization, which leads to certain difficulty during the scheduling of tasks in cloud computing newline