Speedup and scalability analysis of Master–Slave applications on large heterogeneous clusters

Although cluster environments have an enormous potential processing power, real applications that take advantage of this power remain an elusive goal. This is due, in part, to the lack of understanding about the characteristics of the applications best suited for these environments. This paper focuses on Master/Slave applications for large heterogeneous clusters. It defines application, cluster and execution models to derive an analytic expression for the execution time. It defines speedup and derives speedup bounds based on the inherent parallelism of the application and the aggregated computing power of the cluster. The paper derives an analytical expression for efficiency and uses it to define scalability of the algorithm–cluster combination based on the isoefficiency metric. Furthermore, the paper establishes necessary and sufficient conditions for an algorithm–cluster combination to be scalable which are easy to verify and use in practice. Finally, it covers the impact of network contention as the number of processors grow.     

Concurrent common knowledge: defining agreement for asynchronous systems

Summary In this paper we present a new, knowledgetheoretic definition of agreement designed for asynchronous systems. In analogy with common knowledge, it is calledconcurrent common knowledge. Unlike common knowledge, it is a form of agreement that is attainable asynchronously. In defining concurrent common knowledge, we give a logic with new modal operators and a formal semantics, both of which are based on causality and consequently capture only the relevant structure of purely asynchronous systems. We give general conditions by which protocols attain concurrent common knowledge and prove that two simple and efficient protocols do so. We also present several applications of our logic. We show that concurrent common knowledge is a necessary and sufficient condition for the concurrent performance of distributed actions. We also demonstrate the role of knowledge in taking snapshots for stable property detection and asynchronous broadcasts. In general, applications that involve all processes reaching agreement about some porperty of a consistent global state can be understood in terms of concurrent common knowledge. Prakash Panangaden was born in Pune, India in 1954. He attended Calcutta Boys' School and subsequently attended the Indian Institute of Technology, Kanpur where he received an M.Sc. in Physics in 1975. He went to graduate school at the University of Chicago where he studied relativity. He moved to the University of Wisconsin-Milwaukee to study with Leonard Parker to work on quantum field theory on curved space times. After a post-doc at the University of Utah, he decided that it was time for something completely different. He began a Masters with Robert Keller on the semantics of indeterminate dataflow networks. He became an Assistant professor at Cornell University in 1985 and an Associate Professor at McGill University in 1990. He has also made extended visits to the Computer Laboratory, University of Cambridge and to the C.W.I. Amsterdam. Kim Taylor has been an assistant professor at the University of California at Santa Cruz since July 1990. Her research interests are in the design and analysis of algorithms for distributed systems. She received the BS degree in Electrical Engineering and Computer Science from Rice University in May 1985, and the PhD in Computer Science from Cornell University in August 1990.An earlier version of this work appears in Proceedings of the Seventh Annual ACM Symposium on Principles of Distributed Computing, August 1988Supported in part by NSF grants DCR-8602072 and CCR-8818979.Supported in part by an AT&T Ph.D. Scholarship.     

基于DoDAF的武器装备体系需求分析方法

武器装备体系是为提供预期能力而组合在一起的相互依赖的武器装备系统的集合,武器装备体系需求分析是当前装备需求论证的热点和难点问题.在简要分析了武器装备体系需求研究现状的基础上,通过对比结构化分析方法和面向对象分析方法的优缺点,提出综合利用两类方法,基于DoDAF标准进行体系需求组合建模的基本思路和基本步骤,从体系的角度实现作战需求向装备需求的映射,对如何进一步完善体系需求分析方法提出针对性的建议.     

Distributed model predictive control of nonlinear systems subject to asynchronous and delayed measurements

In this work, we design distributed Lyapunov-based model predictive controllers for nonlinear systems that coordinate their actions and take asynchronous measurements and delays explicitly into account. Sufficient conditions under which the proposed distributed control designs guarantee that the state of the closed-loop system is ultimately bounded in a region that contains the origin are provided. The theoretical results are demonstrated through a chemical process example.     

Markov-chain based reliability analysis for distributed systems

In a typical distributed computing system (DCS), nodes consist of processing elements, memory units, shared resources, data files, and programs. For a distributed application, programs and data files are distributed among many processing elements that may exchange data and control information via communication link. The reliability of DCS can be expressed by the analysis of distributed program reliability (DPR) and distributed system reliability (DSR). In this paper, two reliability measures are introduced which are Markov-chain distributed program reliability (MDPR) and Markov-chain distributed system reliability (MDSR) to accurately model the reliability of DCS. A discrete time Markov chain with one absorbing state is constructed for this problem. The transition probability matrix is employed to represent the transition probability from one state to another state in a unit of time. In addition to mathematical method to evaluate the MDPR and MDSR, a simulation result is also presented to prove its correction.     

Efficient distributed snapshots in an anonymous asynchronous message-passing system

We present a global snapshot algorithm with concurrent initiators, with termination detection in an anonymous asynchronous distributed message-passing system having FIFO channels. In anonymous systems, process identifiers are not available and an algorithm cannot use process identifiers in its operation. Such systems arise in several domains due to a variety of reasons. In the proposed snapshot algorithm for anonymous systems, each instance of algorithm initiation is identified by a random number (nonce); however, this is not used as an address in any form of communication. In the algorithm, each process can determine an instant when the local snapshot recordings at all the processes have terminated. This is a challenging problem when an algorithm cannot use process identifiers and a process does not know the number of processes in the system or the diameter of the network and cannot use a predefined topology overlay on the network, because there is no easy way to identify the global termination condition. The message complexity of our algorithm is (cn²)$\left(c{n}^2\right)$, where c$c$ is the number of concurrent initiators and n$n$ is the number of processes in the system, which is much better than that of the algorithm by Chalopin et al. (2012) [6]. Further, the algorithm by Chalopin et al. also requires knowledge of the network diameter.     

A comparative content analysis of face-to-face vs. asynchronous group decision making

A field experiment was conducted to analyze the process and contents of group discussions. Groups solved a case study either orally or through an asynchronous computer-mediated communication system. Findings show that asynchronous groups had broader discussions and submitted more complete reports than their face-to-face counterparts. However, there was no difference in the ability to transfer information from the discussion to the report; under both conditions, about 15% of the issues mentioned in the discussion were omitted from the final report. In terms of coordination, face-to-face teams covered the case study questions sequentially, while asynchronous groups were more focused on solving their general disagreements.     

Event‐driven asynchronous method calls with the D‐Bus message system

The use of inter‐process communication can yield many benefits for event‐driven desktop software. However, inter‐process communication (IPC) research has traditionally been focused on calculation‐oriented distributed computing. This article shows that existing IPC solutions are a poor fit for the constraints imposed by event‐driven programming. Our novel solution is built on top of the D‐Bus system, which enjoys increased practical usage, but is still scantily researched. We focus on efficient handling of asynchronous D‐Bus method calls, in a fashion similar to how Hypertext Transfer Protocol requests are treated in Asynchronous JavaScript and XML Web applications. This is supplemented with two design patterns that simplify processing of results for many kinds of asynchronous operations in event‐driven software, besides just D‐Bus calls. Our approach is shown to be more appropriate for event‐driven applications than traditional remote procedure call systems in aspects as diverse as interactivity, threading complexity and electrical power usage. Copyright © 2013 John Wiley & Sons, Ltd.     

一种异步电动机故障诊断新方法

针对基于RBF神经网络的异步电动机故障诊断方法存在参数确定较困难的问题,提出了一种基于差分进化算法优化RBF神经网络的异步电动机故障诊断方法。首先采用小波变换对异步电动机运行状态信号进行消噪处理,然后采用主元分析法与小波包分析法相结合方式提取消噪后的异步电动机运行状态信号特征,最后采用差分进化算法优化后的RBF神经网络对异步电动机运行状态信号特征进行诊断。实验结果表明,与未优化的RBF神经网络相比,采用差分进化算法优化后的RBF神经网络可有效识别出异步电动机故障。     

Optimization of program recursive function calls analysis method

Static analysis can be classified into intra-procedural analysis and inter-procedural analysis-by-analysis objects. Inter-procedural analysis need to analyze the whole program adopting bottom-up parsing algorithms. Inter-procedural analysis can be converted to intra-procedural analysis by logging the context of follower’s function call in bottom-up analysis. However, the existence of recursive function calls in program increases the difficulty of inter-procedural analysis, which affects the detection accuracy of static analysis to some extent. An optimal method for analyzing the recursive function calls of program is presented in this paper. This method can lift the recursive function calls in function call graph of the program to be analyzed on the premise that vital function calls will be preserved. In this way, the false negative rate and false positive rate of program static analysis can be reduced and the accuracy of program security check can be improved.     

一种异步电动机故障诊断新方法

针对基于RBF神经网络的异步电动机故障诊断方法存在参数确定较困难的问题,提出了一种基于差分进化算法优化RBF神经网络的异步电动机故障诊断方法。首先采用小波变换对异步电动机运行状态信号进行消噪处理,然后采用主元分析法与小波包分析法相结合方式提取消噪后的异步电动机运行状态信号特征,最后采用差分进化算法优化后的RBF神经网络对异步电动机运行状态信号特征进行诊断。实验结果表明,与未优化的RBF神经网络相比,采用差分进化算法优化后的RBF神经网络可有效识别出异步电动机故障。     

Stability of nonlinear asynchronous systems

In this work, we focus on a class of nonlinear asynchronous systems defined by two different modes of operation, one stable and the other one unstable. The switching between the two modes of operation is driven by external asynchronous events. It is assumed that on any time interval of a given length, the maximum time in which the system evolves in the unstable mode is bounded. This property is given in the form of a rate constraint. Under this assumption, we study the behavior of this class of systems and provide existential results of conditions on this rate constraint under which various types of stability of the origin of the nonlinear asynchronous system can be assured.     

Scalability analysis of three monitoring and information systems: MDS2, R-GMA,and Hawkeye

Monitoring and information system (MIS) implementations provide data about available resources and services within a distributed system, or Grid. A comprehensive performance evaluation of an MIS can aid in detecting potential bottlenecks, advise in deployment, and help improve future system development. In this paper, we analyze and compare the performance of three implementations in a quantitative manner: the Globus Toolkit^®${}^{\circledR }$ Monitoring and Discovery Service (MDS2), the European DataGrid Relational Grid Monitoring Architecture (R-GMA), and the Condor project's Hawkeye. We use the NetLogger toolkit to instrument the main service components of each MIS and conduct four sets of experiments to benchmark their scalability with respect to the number of users, the number of resources, and the amount of data collected. Our study provides quantitative measurements comparable across all systems. We also find performance bottlenecks and identify how they relate to the design goals, underlying architectures, and implementation technologies of the corresponding MIS, and we present guidelines for deploying MISs in practice.     

Formal timing analysis of distributed systems

Although a large number of formal methods have been reported in the literature, most of them are applicable only at the initial stages of software development. A major reason for this situation is that those formalisms lack expressiveness to describe the behavior of systems with respect to their underlying configurations. On the other hand, recent experience has shown that the complex nature of distributed systems is conveniently described, constructed and managed in terms of their configuration. In this context, with the twin objectives of accurately modelling the real-timed behavior of distributed systems and supporting the analysis of timing behavior with respect to their underlying configurations, we formulate a logic language called distributed logic (DL). DL is a first-order logic augmented with temporal and spatial modalities. The semantics of DL are based on ideas drawn from both the interleaving and partial order models. In addition to the syntax and semantics of the logic, a formal proof scheme for a distributed programming model is also presented. Finally, use of the proof method is illustrated through the analysis of the real-time properties of a sample problem.     

面向方面的信息系统建模方法及应用研究

随着面向对象软件开发技术的不断发展以及软件规模的不断扩大,系统地分析和设计变得日益复杂。软件开发过程中很难利用现有的编程技术（如面向对象编程、面向过程编程）对横切系统多个模块的关注点进行鉴别、理解和模块化。为解决面向对象开发方法中的耦合性较高、可重用性较低和代码分散等问题,将面向方面的编程思想引人到系统分析和设计阶段,提出了一种面向方面的信息系统建模方法。     

面向对象分析在虚拟实验系统开发中的应用

面向对象方法作为现代软件工程技术的重要部分,在现代信息系统的开发中发挥着积极的作用.面向对象分析是面向对象开发的重要步骤,其主要工作内容是研究问题域和用户需求,运用面向对象的观点和原则发现问题域中与系统责任有关的对象及对象的特征和相互关系,从而最终建立一个完整、确切地反映问题域和用户需求且独立于实现的系统模型.在对基本原理介绍的基础上,重点描述了面向对象分析技术在网上虚拟实验系统开发中的应用.     

A trust-aware,self-organizing system for large-scale federations of utility computing infrastructures

On-demand computing environments, like Cloud/Grid systems, consist of nodes that individually manage local resources intended to be served to clients. When a client needs some resources, it has the problem of finding the most suitable nodes capable of providing them. In addition, a provider node too may be in need to efficiently locate resources for itself, given the emerging, highly competitive, context of large-scale federations. Indeed, a node competes, with the other federated ones, to obtain the assignment of available tasks. To this end, it may decide to publish a set of resources/ services wider than the one it has currently available. Should such a node be assigned a job for which its actual resources are insufficient, it could end up requiring the collaboration of other nodes.Hence the crucial problem, for nodes and clients alike, is to determine the most promising collaborators. For this purpose, in the competitive and demanding scenarios considered, we advocate taking into account the trustworthiness of nodes in declaring their capabilities, i.e., to help it making an effective selection of possible collaborators, each node should be provided with a trust model for accurately evaluating the trustworthiness of its interlocutors.In this paper, a trust-based approach for large-scale federations Utility Computing infrastructures is proposed. The proposed model is designed to allow any node to find the most suitable collaborators in an efficient way, avoiding exploration of the whole node space. A fully decentralized approach is employed, which allows nodes of a federation to be organized in an overlay network on the basis of suitable criteria. This enables any customer or provider in need of collaborators to determine a suitable set of candidate nodes within which to search.     

基于Agent的复杂系统分布仿真建模方法的研究

基于Agent的分布仿真是研究大型复杂系统的一种有效的、重要的方法。为了减小复杂系统仿真的复杂度,增加仿真模型的重用和可维护性,需要研究基于Agent分布仿真的建模方法。首先对复杂系统及其特性进行了分析,对基于Agent的仿真进行了全面的论述,然后对基于Agent的复杂系统仿真中的复杂系统建模分析、Agent建模分析以及Agent的分布进行了分析,给出了基于Agent的复杂系统分布仿真的建模步骤,最后给出了在此建模思想指导下的金融证券市场的建模过程。