Multi-Core Scheduling in Cloud Computing Abhishek Kaushik Inderveer Chana Computer Science and Engineering Department Computer Science and Engineering Department Thapar University Thapar University Patiala-147004 Patiala- 147004 kaushikabhishek87@gmail.com inderveer@thapar.edu Abstract – I.T. industry keeps on looking for high processing power, the reason behind the introduction of multi-core processors. Cloud computing is also to adopting multi-core processors. Prologue of multi-core processors in cloud computing comes with discrepancies, one of which is resource scheduling. The virtual machine that runs on cloud fights for the core in multi-core processors. This leads to problems like starvation and under-utilization. There are several techniques for multi-core resource scheduling but none of present techniques focuses on resource scheduling of multi-core processors especially in cloud environment. A new approach is proposed in this paper which overcomes the problems which occurs when present multi- core resource scheduling techniques compared in the cloud environment. Keywords- cloud, multi-core processors, scheduling I. INTRODUCTION With the development of Computer and Internet, new challenges are arising day by day. One of those challenges is increasing in demand of connectivity and data handling.. To overcome these challenges, Cloud Computing emerges as a elastic infrastructure that can be managed according to increase or decrease in resource demand. As being an I.T technology, it needs to change according to the latest I.T trends to cope up with customer demands. Figure 1: Multi-Core Processors in Cloud Computing One of these trends is Multi-Core processors. Continues increasing demand of “high processing” [1] pushing single- core processors out of market. So, it becomes mandatory for cloud computing to adopt multi-core processors [2] but this adaptation generate whole new set of problem. The issue here is incompatibility of virtualization [3] and multi-core processors. Virtual machines running on cloud always fight for resources. In case of multi-core processors resources are number of cores. Due to this race of virtual machines for getting maximum number of cores, situations like starvation and underutilization occurs. As virtualization and multi-core both are inseparable part of cloud computing, it is necessary that they must work in harmony with each other. Further organization of this paper is as, section second discusses related work. In section three a comparison of existing resource scheduling techniques is done. Section four present the proposed technique. Section five is conclusion and future scope followed by section six references. . II RELATED WORK There are several researchers are working on the resource scheduling in multi-core processors. Some of these are discussed here which are most relevant to the work we are presenting. Zhiyuan Shao,. Jian,Huang, Hai Jin, Kan presents an Underlying Layout Aware Scheduling (ULAS) [4] technique. They proposed mapping of virtual cpus on physical cpu can be done to according to the workload generated on respected vrtual machines. Jian Huang, Hai Jin, Kan Hu, Zhiyuan Shao [5] presents a resource scheduling technique for cases like threshing .They say that the history of resource utilization can be used to allocate the future resources for using a technique Forecasting and Time Delay Subtraction (FTDS). Y. Zhang, A. Bestavros, M. Guirguis, I. Matta, and R. West suggest [6] changing the Virtual Machine itself in such a way that they can share the resources evenly without fighting for resources. Yong Park, Dongkun Shin [7] explains a similar technique that utilizes the history of resource utilization. In their technique guest operating system utilize the history of resource utilization and feed that data to the virtual machine monitor for the proper resource allocation. All these techniques discussed above surely presents the effective method for multi-core resource scheduling but none of these techniques have been implemented in the context of cloud computing. Whereas our goal in this paper is to provide resource allocation technique in case of multi-core processors with respect to the cloud computing.