This volume contains papers presented at The 2013 International Conference on Grid & Cloud Computing & Applications (GCA’13). Their inclusion in this publication does not necessarily constitute endorsements by editors or by the publisher.
Additional Info
  • Publisher: Laxmi Publications
  • Language: English
  • Chapter 1

    VM Migration Approach for Autonomic Fault Tolerance in Cloud Computing Price 2.99  |  2.99 Rewards Points

    VM Migration for Autonomic Fault Tolerance provides the continuous availability for applications in the event of virtual machine failures along with the benefits of improved performance and manageability. Multipath correlation coefficient approach for VM migration has been proposed, which takes into account the overutilization of CPU without performance degradation based on work variability. The proposed approach has been validated through CloudSim toolkit. The experimental results demonstrate that the proposed approach is effective for handling VM migration or VM Reallocation. Further the proposed algorithm considerably reduces the VM Reallocation execution time, VM Reallocation time standard deviation and number of VM migrations.

  • Chapter 2

    Adaptive Data Replication Scheme Based on Access Count Prediction in Hadoop Price 2.99  |  2.99 Rewards Points

    Hadoop, an open source implementation of the MapReduce framework, has been widely used for processing massive-scale data in parallel. Since Hadoop uses a distributed file system, called HDFS, the data locality problem often happens (i.e., a data block should be copied to the processing node when a processing node does not possess the data block in its local storage), and this problem leads to the decrease in performance. In this paper, we present an Adaptive Data Replication scheme based on Access count Prediction (ADRAP) in a Hadoop framework to address the data locality problem. The proposed data replication scheme predicts the next access count of data files using Lagrange’s interpolation with the previous data access count. With the predicted data access count, our adaptive data replication scheme determines whether it generates a new replica or it uses the loaded data as cache selectively, optimizing the replication factor. Furthermore, we provide a replica placement algorithm to improve data locality effectively. Performance evaluations show that our adaptive data replication scheme reduces the task completion time in the map phase by 9.6% on average, compared to the default data replication setting in Hadoop. With regard to data locality, our scheme offers the increase of node locality by 6.1% and the decrease of rack and rack-off locality by 45.6% and 56.5%, respectively.
     

  • Chapter 3

    Elastic Integration of a Dedicated Grid Site with an Opportunistic Cloud Infrastructure Price 2.99  |  2.99 Rewards Points

    This work aims to extend and integrate a dedicated grid infrastructure based on the middleware gLite with opportunistic IaaS resources provided by the UnaCloud project. The extension takes place during the periods of time when the current demand exceeds the base resources capacity. The integration is done elastically, allowing the grid site to dynamically change the processing capabilities according to the current workload, using opportunistic resources when needed. To manage the elastic integration different implementation strategies are proposed and different testing scenarios are analyzed to validate a good reactiveness to workload spikes.

  • Chapter 4

    Instantaneous Availability-Aware Scheduling Method for Parallel Task in Cloud Environment Price 2.99  |  2.99 Rewards Points

    Cloud computing is a new emerging computing paradigm that advocates supplying users everything as a service. High availability is a key requirement in design and development of cloud computing systems where processors operate at different speeds and are not continuously available for computation. This paper addresses the main factors of availability requirement for parallel tasks based on the graph structure of parallel task. We present the formulas to quantify the availability requirements of parallel tasks and availability support of computing resources. Through being aware of availability, we realize the availability matching between availability requirement and availability support, and develop an availability aware scheduling algorithms, that is Availability-DLS.

  • Chapter 5

    BLASTer: A Hub-based Tool for Bioinformatics Price 2.99  |  2.99 Rewards Points

    Basic Local Alignment Search Tool (BLAST) is a widely used sequencing tool in the bioinformatics community. BLAST searches are highly-parallelizable, lending themselves to computation on a high-throughput resource. In this paper, we describe the BLASTer tool developed by Purdue University to present an easy-to-use, web-based interface for performing BLAST searches. BLASTer uses the large HTCondor pool at Purdue University to dramatically shorten run times compared to desktop usage. Using the HUBzero platform, BLASTer can be made available for research and instruction to users worldwide. This paper describes the platform, the development process, early success stories, and efforts to grow the user community.
     

  • Chapter 6

    An Efficient Method by Using Prediction to Reduce Times in Live Migration of Virtual Machine Price 2.99  |  2.99 Rewards Points

    Cloud system, especially IaaS has became hot issue because of its low TCO. In the field of IaaS, server virtualization has close connection due to benefits of resource sharing, fault tolerance, portability and cost efficiency. Live VM migration is one of technologies which manage virtualized server. Although live VM migration just moves virtual machine to another within few downtime, it can cause more network traffic because of longer migration time. So we suggest a method to reduce migration time through predicting iterations of pre-copy phase. Our method determines number of iterations using average dirty page rate and decides stop condition. We apply this method to Xen and demonstrate our method works efficiently.

  • Chapter 7

    Towards a User Deployable Service-oriented Autonomic Multi-cloud Overlay Infrastructure for Sky Computing Price 2.99  |  2.99 Rewards Points

    Cloud computing services at the SaaS level of the cloud computing model are easy to access and utilize. In contrast, the self-service approach taken at the IaaS level results in difficulties for non-savvy users. Compounding these difficulties is the fact that for resiliency, economic, and scalability reasons, utilization of IaaS resources across multiple cloud providers in a unified manner is deemed the best strategy. In this paper, we outline our proposal for a user-centric multi-cloud autonomic overlay infrastructure that can be deployed across existing cloud systems without the need for specialized hardware or action on the part of cloud providers. We also present the results of a simple genome sequencing experiment and bandwidth measurements conducted on a crude prototype implementation of the system.

  • Chapter 8

    SLA-Aware Adaptive Provisioning Method Price 2.99  |  2.99 Rewards Points

    Resource provisioning is a general technique for handling the resource allocation in cloud environment. Monitoring the system performance and the user request is crucial for efficient cloud resource management. Also in cloud environments, issues such as cost and resource provisioning based on QoS constraints are yet to be addressed. In this paper,we present a SLA(Service Level Agreement) - Aware Adaptive (SAA) provisioning method for Virtualized Applications that employ a flexible determining model. We present advanced cloud infrastructure which maintain proper virtual machine numbers by optimizing resources allocation. Our experiments show that our adaptive model minimize the maintain cost of virtual machines while satisfying user average response time constraint and the request arrival rate constraint.

  • Chapter 9

    Harnessing the Benefits of Big Data in the Cloud Price 2.99  |  2.99 Rewards Points

    In recent years, data has gained more visibility and importance to organizations and governments worldwide. Innovative technologies have emerged to more efficiently collect, process and disseminate data or information, bringing transformational value to these organizations. In the current information age, the transformation of data to useful information is playing a vital role in driving effectiveness and efficiency. Innovation in social computing, the proliferation of mobile devices, and the emergence of cloud computing have magnified data availability and accessibility. These changes have caused a shift from traditional data management systems and processes to an emerging paradigm of Big Data. In this paper we present a Cloud based Big Data platform for collecting and transforming distributed data into knowledge or insightful information.

  • Chapter 10

    A Stream Symmetric Homomorphic Encryption Scheme with Finite Price 2.99  |  2.99 Rewards Points

    This paper presents a new homomorphic encryption scheme which uses sequential machines as a basis to produce a stream-symmetric encryption, which can be expanded to a homomorphic encryption with the help of the operation over automata called serial composition, allowing the application of any operation that can be represented by means of a finite automaton, on the set which contains the cipher-text. Unlike conventional cryptographic schemes this is not aimed at sharing information securely, so key distribution
    is not necessary and a symmetrical or asymmetrical cryptosystem makes no difference. With regard to safety, the intruder can only design cryptanalytic algorithms having as input a set of encrypted text and a composite automaton, constructed with the keys and an operation automaton. To illustrate the correct operation of the cryptographic scheme, we include the implementation of the greatest common divisor of two numbers.
     

  • Chapter 11

    Energy Efficiency Models Implemented in a Cloud Computing Environment. Price 2.99  |  2.99 Rewards Points

    Cloud computing is a paradigm of computing that shifts the way computing has been done in the past. In cloud computing, recourses are rented. For example cloud computing provides flexible services to cloud users services such as software, platform and infrastructure as services.

  • Chapter 12

    A system of Cloud Services and SOA to improve Health Care Organizations Price 2.99  |  2.99 Rewards Points

    The interoperability among different systems in healthcare is an historical problem. The development of
    interoperability standards, the interest for an approach based on Personal Health Record (PHR) and Electronic Health Record (HER) led a significant evolution in the field of the ehealth. In the meanwhile, Cloud Computing (CC) is a very interesting solution for IT Applications. This paradigm presents many advantages for management.
     

  • Chapter 13

    Essential Cloud Security Features in Windows Azure Price 2.99  |  2.99 Rewards Points

    Cloud computing technology is recently gaining widespread popularity among business owners and consumers/users for hosting and delivering services over the Internet. This technology offers users on-demand access to shared resources, services, and applications with the Internet access by eliminating the need for tedious installation procedures. Security and privacy issues in cloud computing is one of the major barriers for the wide adoption of this emerging technology. In this paper, we first discuss the security and privacy guidelines pertinent to the public cloud computing environment as described by NIST. We then investigate different security features of the Microsoft Azure cloud computing platform and analyze how the security and privacy guidelines described by NIST are implemented in this cloud platform.
     

  • Chapter 14

    An Approach for QoS-aware Cloud Service Selection Based on Genetic Algorithm and Simplex Method Price 2.99  |  2.99 Rewards Points

    For the problem of cloud service selection, this paper gives a hybrid algorithm. The hybrid algorithm is composed mainly by Simplex Method and Genetic Algorithm. In this algorithm, a tree traversal sequence encoding scheme and some Simplex Method operations are proposed. The design of the tree traversal sequence encoding can support various types of service combinations. In addition, the hybrid algorithm uses Simplex Method operations to further improve local convergence of Genetic Algorithm.

  • Chapter 15

    Maximizing SLA and QoE in Heterogeneous Cloud Computing Environment Price 2.99  |  2.99 Rewards Points

    Cloud Computing delivers users a proficient way to dynamically allocate computing resources to meet demands. The use of Server virtualization techniques for Cloud Computing platforms provide great elasticity with the capability to consolidate several virtual machines on the same physical server, to resize a virtual machine capacity and to migrate virtual machine across physical servers. A key challenge for Cloud Service providers is proper resource management while taking into account of Service Level Agreement and Quality of Experience of users. Reserving resources would be beneficial to improve the quality of service, if the actual demand of the user is known in advance.

  • Chapter 16

    Social Network Storage Allocation Price 2.99  |  2.99 Rewards Points

    Today, social networks present massive amounts of data by the hour that need storage, therefore, along with the aid of cloud computing, social networks users can have their data stored in data centers anywhere around the globe belonging to the cloud. This paper will be focusing on how to allocate user data to the appropriate global data centers from a social networking point of view.

  • Chapter 17

    A Fast Live Migration Algorithm of Virtual Machine with CPU Scheduling Price 2.99  |  2.99 Rewards Points

    Live migration of virtual machine (VM) is one of the most important virtualization technologies to facilitate the management and improve the performance of data centers. In this paper, we propose a fast convergent live migration approach called FMC. It is a self-adaptive method which synchronizes the state of the VM migrated between the source and target host by log tracing and replaying instead of dirty memory pages transferring.

  • Chapter 18

    A Fuzzy Predictable Load Balancing Approach in Cloud Computing Price 2.99  |  2.99 Rewards Points

    Cloud computing is a new paradigm for hosting and delivering services on demand over the internet where users access services. It is an example of an ultimately virtualized system, and a natural evolution for data centers that employ automated systems management, workload balancing, and virtualization technologies. Live virtual machine (VM) migration is a technique to achieve load  balancing in cloud environment by transferring an active overload VM from one physical host to another one without disrupting the VM.

About the Author

Professor of Computer Science view complete profile

Ashu M. G. Solo is an independent interdisciplinary researcher and developer, electrical engineer, computer engineer, intelligent systems engineer, political and public policy engineer, mathematician, political writer, public policy analyst, political operative, entrepreneur, former infantry platoon commander understudy, and progressive activist. Solo has over 500 research and political commentary publications. view complete profile

Other Books by Author