Efficient Optimal Algorithm of Task Scheduling in Cloud Computing Environment (1404.2076v1)

Abstract: Cloud computing is an emerging technology in distributed computing which facilitates pay per model as per user demand and requirement.Cloud consist of a collection of virtual machine which includes both computational and storage facility. The primary aim of cloud computing is to provide efficient access to remote and geographically distributed resources. Cloud is developing day by day and faces many challenges, one of them is scheduling. Scheduling refers to a set of policies to control the order of work to be performed by a computer system. A good scheduler adapts its scheduling strategy according to the changing environment and the type of task. In this research paper we presented a Generalized Priority algorithm for efficient execution of task and comparison with FCFS and Round Robin Scheduling. Algorithm should be tested in cloud Sim toolkit and result shows that it gives better performance compared to other traditional scheduling algorithm.

• The paper introduces a Generalized Priority Algorithm (GPA) that prioritizes tasks by size and VMs by MIPS to enhance cloud task scheduling efficiency.
• Using CloudSim, the study compared the GPA against FCFS and Round Robin algorithms under varying workloads and tasks.
• Experimental results show the GPA significantly outperforms FCFS and Round Robin in task execution time and resource utilization.

Overview of the Task Scheduling Algorithm in Cloud Computing

The paper "Efficient Optimal Algorithm of Task Scheduling in Cloud Computing Environment" by Dr. Amit Agarwal and Saloni Jain presents a compelling investigation into cloud computing task scheduling methodologies with an emphasis on efficiency. The authors address the inherent complexities of cloud computing, particularly within the field of task scheduling. This paper is situated within a larger context of distributed computing systems, where the aim is often to achieve high-performance computing and optimal resource utilization.

Key Contributions

This paper introduces a Generalized Priority Algorithm (GPA) for task scheduling and performs a comparative evaluation against conventional scheduling algorithms such as First Come First Serve (FCFS) and Round Robin (RR). The GPA offers a novel approach to prioritizing computational tasks based on their size and available resources, enhancing the efficiency and throughput of cloud services. The simulation results, conducted on a CloudSim environment, feature both FCFS and RR as benchmarks to validate the efficacy of the proposed algorithm.

Methodology

The authors thoroughly describe the scheduling process in cloud environments which can be generalized into three main stages:

1. Resource Discovering and Filtering: Identification and status gathering of resources available in the network.
2. Resource Selection: Target resources are chosen based on specific parameters related to both tasks and available resources.
3. Task Submission: Assigning tasks to the selected resources for execution.

Within the GPA, tasks are prioritized according to their size, and virtual machines (VMs) are ranked by their MIPS values, allowing for an optimized allocation strategy. The implementation leverages the computational capabilities of VMs where the machine with the highest MIPS receives tasks with the highest computational demand.

Experimental Evaluation

The experimentations were conducted using an Intel i5 2.6 GHz processor on a Windows 7 platform utilizing CloudSim 3.0.3. The experimental setup involved creating five VMs, maintaining uniform RAM sizes, but varying MIPS ratings. Cloudlets, representing the tasks, were prioritized and scheduled on these VMs. The evaluation involved testing the performance of 100 to 500 tasks, analyzing execution times and resource utilization across the FCFS, RR, and GPA scheduling algorithms.

Findings

The experimental results exhibit that the Generalized Priority Algorithm outperforms both FCFS and RR in terms of execution time and resource utilization. The algorithm demonstrates significant improvements in scheduling efficiency, particularly under conditions of varying workloads and task complexity. The paper underscores the effectiveness of the GPA by offering detailed performance metrics and analysis to substantiate these findings.

Implications and Future Research

This paper advances the existing body of knowledge by introducing an algorithm tailored to the dynamic conditions of cloud computing environments where resource allocation efficiency is paramount. Its contributions are particularly relevant for environments requiring the optimization of computational tasks across distributed resources.

Potential directions for future research include adapting the Generalized Priority Algorithm for larger, more complex datasets and exploring its applicability within grid environments. Further work could also investigate the algorithm's scalability and how it might handle additional parameters like cost and energy consumption.

In conclusion, this paper provides a crucial step forward in task scheduling strategies for cloud computing, laying the groundwork for further technological advancements and applications in the field of distributed computing.