محیط رایانش ابری و الگوریتم های زمانبندی وظایف با عامل چندگانه

ABSTRACT Task Scheduling Algorithms with Multiple Factor in Cloud Computing Environment

In these days all the application runs via virtual machines and resources are allocated to all virtual machines. All applications are independent, unique in technology and has no connection to each other. Resources are compromised on every event or activity performed with all individual units of products and service.

The concept of scheduling is performing very important role in cloud computing with requirement of users in the market. ‘Mapping the tasks’ is the basic concept of scheduling. This is a necessary condition in successful working of cloud as many factors must be examined for useful scheduling. The feasible resources should be appropriate for execution in task scheduling (Tables 1, 2, 3 and 4). The mechanism of task scheduling can not only satisfy to the user, but also increase the utilization for resources [1]. Load balancing factor must be calculated to acquire more resource utilization. The process of load balancing is distributing the load between various nodes to enhance utilization of resource and task response time while also neglecting the condition where nodes are fully occupied while many other distinct nodes are free or performing for limited work. Load balancing assure that processors in the setup or all node in the rooted network connection does approximately the uniform amount of execution at any present of time. To get the complete cost of every user’s applications, all individual service of resources (like Processor cost, Internal memory used, Input/Output cost, etc.) need to be calculated. When the complete cost of all resources has been identified, factual cost and output dissection depend on it can be retrieving, related to all of the traditional concepts of scheduling. Traditional concept leading the use of absolute tasks for users and exceeds the overheads in applications of cloud computation. It may be true that any distinct tasks may not the reason of exceeding costs for Table 1 FCFS Vs VM-Tree

Table 2 FCFS Vs PSO

Table 3 QoS of FCFS

Table 4 FCFS Vs ABC

resources in traditional way. The result is that exceeds in estimate and raises the cost. To compete towards market, some organization has had to reduce the cost or prices of expansive items or products. But they have been capable to get huge mark-ups on less expensive tasks. Minimization in cost or price has raised marketing for individual units/item but degrades the complete strength and good mark-ups on specific tasks or product didn’t equal the down in the confine output of expensive items [2]. However, Load Balancing and Allocation Cost are the primary issues for task scheduling algorithms of user’s applications in cloud computation. Presently the complete cost and proper resource utilization of the scheduling issue has encouraged researchers to recommend multifarious cost related task scheduling algorithms. More advanced algorithm designed on these factors are introduced by innovators or researchers such as Activity Based Cost, Particle Swarm Optimization, DLA (Double Level Priority) and Balancing the Load [1–3, 6–13] etc. This paper proposes an implementation view for scheduling techniques with simulated outcome in CloudSim 3.0 simulator by taking the factor cost and balancing the load. The remaining part of this research paper is implemented as followed: Sect. 2 explains/methodology. Section 3 derives the implementation by simulation and analysis. Section 4 shows the conclusion.