Heterogeneous networked clusters are being increasingly used as platforms for resource-intensive parallel and distributed applications. The fundamental underlying idea is to provide large amounts of processing capacit...
Heterogeneous networked clusters are being increasingly used as platforms for resource-intensive parallel and distributed applications. The fundamental underlying idea is to provide large amounts of processing capacity over extended periods of time by harnessing the idle and available resources on the network in an opportunistic manner. In this paper we present the design, implementation and evaluation of a framework that uses JavaSpaces to support this type of opportunistic adaptive parallel/distributed computing over networked clusters in a non-intrusive manner. The framework targets applications exhibiting coarse grained parallelism and has three key features: (1) portability across heterogeneous platforms, (2) minimal configuration overheads for participating nodes, and (3) automated system state monitoring (using SNMP) to ensure non-intrusive behavior. Experimental results presented in this paper demonstrate that for applications that can be broken into coarse-grained, relatively independent tasks, the opportunistic adaptive parallel computing framework can provide performance gains. Furthermore, the results indicate that monitoring and reacting to the current system state minimizes the intrusiveness of the framework.
It is a well-known fact that software documentation is, in practice, poor and incomplete. Though specification, design, and test documents-among other things-are required by standards and capability maturity models (e...
详细信息
It is a well-known fact that software documentation is, in practice, poor and incomplete. Though specification, design, and test documents-among other things-are required by standards and capability maturity models (e.g., SEI CMM), such documentation does not exist in a complete and consistent form in most organizations. When documents are produced, they tend to follow no defined standard and lack information that is crucial to make them understandable and usable by developers and maintainers. Then a fundamental practical question, which motivated this keynote address, is to better understand what type of documentation is required, what is needed to support its completeness and consistency, and what is the level of precision required for each type of document. These questions cannot be investigated at that level of generality though. Answers are likely to be very context-dependent if they are to be precise. We focus our attention here on object-oriented development and the Unified Modeling Language (UML).
With the rise of distributed e-commerce in recent years, demand for automated negotiation has increased. In turn, this has facilitated a demand for ever more complex algorithms to conduct these negotiations. As the co...
详细信息
With the rise of distributed e-commerce in recent years, demand for automated negotiation has increased. In turn, this has facilitated a demand for ever more complex algorithms to conduct these negotiations. As the complexity of these algorithms increases, our ability to reason about and predict their behaviour in an ever larger and more diverse negotiation environment decreases. In addition, with the proliferation of internet-based negotiation, any algorithm also has to contend with potential reliability issues in the underlying message-passing infrastructure. These factors can create problems for building these algorithms, which need to incorporate methods for survival as well as negotiation. This paper proposes a simple yet effective framework for integrating survivability into negotiators, so they are better able to withstand imperfections in their environment. Results of an experiment are provided which show how the stability of a negotiation community is affected by incorporating an example survival behaviour into negotiators operating in an environment developed to support this framework.
The purpose of this paper is to propose visual models for a Web application using Java and XML related technologies. We consider a Web application that uses 3-tier architecture and each tier is distributed onto Web, a...
详细信息
The ability to efficiently discover information using partial knowledge (for example keywords, attributes or ranges) is important in large, decentralized, resource sharing distributed environments such as computationa...
详细信息
ISBN:
(纸本)9780769519654
The ability to efficiently discover information using partial knowledge (for example keywords, attributes or ranges) is important in large, decentralized, resource sharing distributed environments such as computational grids and peer-to-peer (P2P) storage and retrieval systems. This paper presents a P2P information discovery system that supports flexible queries using partial keywords and wildcards, and range queries. It guarantees that all existing data elements that match a query are found with bounded costs in terms of number of messages and number of peers involved. The key innovation is a dimension reducing indexing scheme that effectively maps the multidimensional information space to physical peers. The design, implementation and experimental evaluation of the system are presented.
This article is a first attempt towards a comprehensive, systematic methodology for class interface testing in the context of client/server relationships. The proposed approach builds on and combines existing techniqu...
详细信息
This article is a first attempt towards a comprehensive, systematic methodology for class interface testing in the context of client/server relationships. The proposed approach builds on and combines existing techniques. It first consists in selecting a subset of the method sequences defined for the class testing of the client class, based on an analysis of the interactions between the client and the server methods. Coupling information is then used to determine the conditions, i.e., values for parameters and data members, under which the selected client method sequences are to be executed so as to exercise the interaction. The approach is illustrated by means of an abstract example and its cost-effectiveness is evaluated through a case study.
We previously developed a framework, Hydra, for adding formal semantics to a collection of UML diagrams that enable the automated derivation of formal language specifications for those diagrams. Recently, we have also...
详细信息
The distributed telecommunications sector not only requires minimal time to market, but also software that is reliable, available, maintainable and scalable. High level programming languages have the potential to redu...
详细信息
ISBN:
(纸本)1581137729
The distributed telecommunications sector not only requires minimal time to market, but also software that is reliable, available, maintainable and scalable. High level programming languages have the potential to reduce development time and improve maintainability due to their compact code size. Moreover reliability is improved by safe type systems and relatively easy verification. This paper outlines plans and initial results from a joint project between Motorola and Heriot-Watt University that aims to evaluate the suitability of distributed functional languages for constructing telecommunications software. The evaluation will use the ERLANG and Glasgow distributed Haskell(GdH) languages, and be based on the construction of several typical applications. The evaluation will focus on reliability issues like ease of verification, availability issues like fault-tolerance or resilience, as well as whether the languages deliver the required functionalities, like real-time capabilities. The impact of specific languages techniques will also be assessed, including type system, strictness, validation and distributed coordination. The ERLANG and GdH implementations of the applications will be compared with existing C++/CORBA and Java/JINI implementations. The first application, a Dispatch Call Controller(DCC), has been constructed in ERLANG and measured on a Beowulf cluster. We find that the DCC scales, achieving a relative speedup of 14.5 on 16 processors. The DCC is resilient, achieving 105% throughput at 200% load and 56% throughput at 9000% load on 16 processors. The DCC is fault-tolerant, remaining available despite any one process or processor failure. The DCC has dynamic adaptability, remaining available as processors are added or removed. Copyright 2003 ACM.
In this paper we present the design, implementation and evaluation of the Grid-enabled Discover middleware substrate. The middleware substrate enables Grid infrastructure services provided by the Globus Toolkit (secur...
详细信息
暂无评论