Adaptive Distributed Admission Control in Differentiated Services Domains

In this paper we propose a scalable admission control scheme for the QoS sensitivity traffic in DiffServ domains. In our scheme, the ingress touters perform admissibility test in a fully distributed and parallel fashion for requests based on our resource per-assigning mechanism. Then, we introduce a novel two phase token passing mechanism to adaptively optimize resource per-assigning among contending edge touters in proportion to their traffic. In addition, we adopt a measurement based admission decision-making criterion to gain the benefit of high utilization of statistical multiplexing. Our simulation results indicate that even under very high request load it is possible to perform admission control and resource allocation in parallel without suffering in terms of response time, packet loss rate, or utilization.

Grid-based internet worm behavior simulator

The traditional network simulator has function and performance limitation when simulating Internet worms,so we designed the grid-based Internet worm behavior simulator (IWBS Grid).IWBS Grid makes use of the real Internet topology,link and routing information,and simulates the worm behavior at the packet event-driven level;and proposes a high-performance Internet worms behavior simulation platform by right of the grid computing capability,resource and task management,and so on.The experimental results show that IWBS grid surpasses the traditional simulator in simulating capability,and the technology to track the worm propagation in packet level can propose the valuable information for the further study on worms.

A Survey of Data Partitioning and Sampling Methods to Support Big Data Analysis

Computer clusters with the shared-nothing architecture are the major computing platforms for big data processing and analysis.In cluster computing,data partitioning and sampling are two fundamental strategies to speed up the computation of big data and increase scalability.In this paper,we present a comprehensive survey of the methods and techniques of data partitioning and sampling with respect to big data processing and analysis.We start with an overview of the mainstream big data frameworks on Hadoop clusters.The basic methods of data partitioning are then discussed including three classical horizontal partitioning schemes:range,hash,and random partitioning.Data partitioning on Hadoop clusters is also discussed with a summary of new strategies for big data partitioning,including the new Random Sample Partition(RSP)distributed model.The classical methods of data sampling are then investigated,including simple random sampling,stratified sampling,and reservoir sampling.Two common methods of big data sampling on computing clusters are also discussed:record-level sampling and blocklevel sampling.Record-level sampling is not as efficient as block-level sampling on big distributed data.On the other hand,block-level sampling on data blocks generated with the classical data partitioning methods does not necessarily produce good representative samples for approximate computing of big data.In this survey,we also summarize the prevailing strategies and related work on sampling-based approximation on Hadoop clusters.We believe that data partitioning and sampling should be considered together to build approximate cluster computing frameworks that are reliable in both the computational and statistical respects.

A Non-Blocking Locking Method and Performance Evaluation on Network of Workstations

network of workstation (NOW) can act as a single and scalable powerful computer by building a parallel and distributed computing platformon top of it. WAKASHI is such a platform system that supports persistent objectmanagement and makes full use of resources of NOW for high performance transaction processing. One of the main difficulties to overcome is the bottleneck causedby concurrency control mechanism. Therefore, a non-blocking locking method isdesigned, by adopting several novel techniques to make it outperform the other typical locking methods such as 2PL: 1) an SDG (Semantic Dependency Graph) basednon-blocking locking protocol for fast transaction scheduling; 2) a massively virtualmemory based backup-page undo algorithm for fast restart; and 3) a multi-processorand multi-thread based transaction manager for fast execution. The new mechanismshave been implemented in WAKASHI and the performance comparison experimentswith 2PL and DWDL have been done. The results show that the new method canoutperform 2PL and DWDL under certain conditions. This is meaningful for Choosing effective concurrency control mechanisms for improving transaction- processingperformance in NOW environments.