International Journal of Computer Trends and Technology (IJCTT) – volume 4 Issue10 – Oct 2013 ISSN: 2231-2803 http://www.ijcttjournal.org Page3507 Secure Data Dynamics in Tandem with Dynamic Resource Allocation Bindu Madhavi #1 , G.Kalpana *2 Dr.R.V.Krishnaiah *3 #1 M.Tech, CSE, DRKIST, Hyderabad, Andhra Pradesh, India # Associate Professor, Department of CSE, DRKIST, Hyderabad, Andhra Pradesh, India # Principal, Department of CSE, DRKIST, Hyderabad, Andhra Pradesh, India Abstract--In Cloud computing both security with perfect data dynamics and optimal resource allocation are essential. For best realization of cloud computing parallel and reliable data processing is required. There are many providers of cloud services such as Oracle, Microsoft, IBM, and Google. The existing systems used for cloud computing are homogenous in nature. The resource allocation and execution of jobs parallelly has some limitations. The security is also concerns as the cloud servers are treated as un-trusted by the cloud users. In this paper parallel processing, dynamic resource allocation challenges are addressed. We built a prototype application to demonstrate the proof of concept and the empirical results are encouraging. Index Terms –Parallel processing, cloud computing, Map Reduce, many-task computing. I. INTRODUCTION Many organizations in the real world are into processing of large volumes of data. This has to be done is a cost effective fashion. Such organizations include Microsoft, Yahoo, and Google etc. They deal with increased volumes of data every day. For this purpose storing and retrieving data using conventional databases is very expensive [1]. To overcome the problem, many companies have started using commodity servers in a big way. When number of such servers is being used for processing huge amount of data, the processing work is divided into multiple tasks and assigned dynamically to every server involved. This facilitates the system to work faster as the available nodes share the job of processing such voluminous data in a short span of time. This is possible due to parallel processing of data. To enhance the performance further these companies have developed customized frameworks that take care of parallel data processing in an efficient fashion. There are many such frameworks existing in the real world. They are created by industry giants such as Microsoft, Yahoo, and Googleetc. For instance the framework developed by Google is known as MapReduce [2]. Microsoft developed a framework known as Dryad [3] while Yahoo developed a framework known as Map- Reduce-Merge [4]. These products are varying capabilities for achieving many task computing or high throughput computing in terms of amount of data to be processed and also the number of tasks involved in the processing [5]. These applications are having different architectural designs. However, they share a common approach in fulfilling objectives such as fault tolerance, parallel programming, optimized executions, and hiding the unnecessary things. These products are basically meant for parallel processing. Developers can write the programs sequentially. When they are given to these products, they are distributed among multiple nodes and the work is done parallel. It is not recommended to establish costly data centers for parallel processing of data when a company involves in processing large volumes of data occasionally. For such organizations, the recommended solution is the usage of cloud computing. Cloud computing has emerged as a technology that enables individuals and organizations to gain access to state-of-the-art servers, data centers, network infrastructure without the need for capital investment in pay per use fashion. Many companies are providing cloud services. They include IBM, Oracle, Microsoft, Google etc. These companies are cloud service providers who provide different kind of cloud services. The services are categorized into IaaS (Infrastructure as a Service), PaaS (Platform as a Service), and SaaS (Software as a service). This paper deals with opportunities and challenges of parallel data processing in IaaS clouds. The cloud products available are EC2 from Amazon [6], Azure from Microsoft etc. The cloud computing technology is based on the concept of virtualization. The virtualasation technology makes the cloud computing a reality as it can reduce the cost of maintaining clouds substantially. The virtualization technology involves creation and destroying of VMs of different