Java has in-built support for distributed computing. As the first language designed from the bottom up with networking and distributed computing in mind, Java makes it easy for computers to cooperate. Students of dist...
详细信息
ISBN:
(纸本)1932415262
Java has in-built support for distributed computing. As the first language designed from the bottom up with networking and distributed computing in mind, Java makes it easy for computers to cooperate. Students of distributed computing need adequate exposure to this technology and in order to meet this need the author has put together a course on distributed computing which uses Java's support for object-oriented programming, clientserver applications, multithreaded servers, security, and distributed objects. This paper describes this course and the programming projects developed and used in the course. Students developed programming projects using open source, Java based middleware technologies that built on the lectures. The focus of the course and projects was providing students with a clear understanding of the trade-offs involved with the different Java based middleware technologies as they design and implement distributedapplications.
It is important for almost aU transaction processing and computer-communication systems to satisfy response time quantile targets. This paper describes HYDRA, a scalable parallel tool for the analytical determination ...
详细信息
ISBN:
(纸本)1892512416
It is important for almost aU transaction processing and computer-communication systems to satisfy response time quantile targets. This paper describes HYDRA, a scalable parallel tool for the analytical determination of response time densities in large, structurally-unrestricted Markov models derived from high-level specifications. The tool exploits an efficient distributed uniformization-based algorithm, combined with hypergraph partitioning to balance computational load across processors while minimising communication. We demonstrate our tool on a 1.6 million state Generalized Stochastic Petri Net model of a flexible manufacturing system, comparing the accuracy of our results with simulation and contrasting the run-time performance of our technique with an approach based on numerical Laplace transform inversion.
We expand the scope of peer-to-peer (P2P) systems to include the concept of "communities". Communities are analogous to interest groups and can overlap. We use communities as a more natural arrangement Of di...
详细信息
ISBN:
(纸本)1892512459
We expand the scope of peer-to-peer (P2P) systems to include the concept of "communities". Communities are analogous to interest groups and can overlap. We use communities as a more natural arrangement Of distributed systems. This paper focuses on a novel, push-pull gossiping technique to improve decentralized information dissemination and search within a P2P community. To facilitate an efficient gossiping operation, a distributed discovery algorithm first identifies highly influential peers (called seers) in a community. Then, the push phase multicasts information to these seers so that peers can easily and quickly retrieve this information via a pull phase. Our experiments show that pushing gossip information to only a small number of seers allows a large percentage of peer members to obtain (pull) the information within just two hops.
The use of object-oriented programming techniques in the development of parallel, finite element analysis software enhances code reuse and makes application development more efficient. In this paper, an object-oriente...
详细信息
ISBN:
(纸本)1892512416
The use of object-oriented programming techniques in the development of parallel, finite element analysis software enhances code reuse and makes application development more efficient. In this paper, an object-oriented programming framework developed by the authors is utilized in the implementation of a parallel finite element software for modeling of the resin transfer molding manufacturing process. The motivation for choosing the resin transfer molding finite element application and implementing it with the object-oriented framework is that it was originally developed and parallelized in a functional programming paradigm. Discussion of the software development effort and performance results are presented and analyzed.
distributed active storage architectures are designed to offload user-level processing to the peripheral from the host servers. In this paper, the author reports preliminary investigation on performance and fault reco...
详细信息
ISBN:
(纸本)9781932415582
distributed active storage architectures are designed to offload user-level processing to the peripheral from the host servers. In this paper, the author reports preliminary investigation on performance and fault recovery designs, as impacted by emerging storage interconnect protocols and novel state-of-the-art storage devices. Empirical results obtained using validated device-level and interconnect data demonstrate the significance of the said parameters on the overall system performance and reliability.
parallel programming has remained a topic of extensive research over the last few decades. With the advent of cluster computing, significant improvements in runtime performance can be achieved by parallelizing sequent...
详细信息
ISBN:
(纸本)1932415262
parallel programming has remained a topic of extensive research over the last few decades. With the advent of cluster computing, significant improvements in runtime performance can be achieved by parallelizing sequential code. The process of parallelizing a serial program is however a labor-intensive and complex operation that involves limitations. This paper describes a software toolkit designed to assist the process of parallelizing sequential code by targeting those functions, subroutines, and loop computations that consume a major portion of the execution time. These time-consuming portions of the code are identified using static analysis as opposed to a more expensive dynamic analysis. The tool is designed to target sequential programs that are written in Fortran. In addition to other features, the tool includes a user-friendly interactive environment that gives feedback to the user regarding the potential bottlenecks in the sequential code. The toolkit has been designed using Perl, Tcl/Tk and uses Message Passing Interface (MPI) as the basis for interprocessor communication to maximize portability.
Most current cycle stealing systems, especially those designed for the Internet require programmers to adopt the master-worker parallel programming paradigm. The master-worker paradigm is extremely limiting from an ab...
详细信息
ISBN:
(纸本)1892512459
Most current cycle stealing systems, especially those designed for the Internet require programmers to adopt the master-worker parallel programming paradigm. The master-worker paradigm is extremely limiting from an abstraction point of view as it requires a single master thread to expose all of the parallelism that the application hopes to exploit. In this paper we describe how we have extended our Internet cycle stealing system G2, to support more general parallel programming paradigms in which any task can create subtasks and in which tasks become self contained units of abstraction. The major problem addressed is how to allow tasks to wait for the results of their subtasks without tying up the resources of volunteer machines. By drawing a programming analogy with the asynchronous implementation of web services in *** we create a familiar yet powerful programming model for creating parallel programs using a variety of paradigms.
This paper looks at the performance behaviors of a methodology for simulating hard sphere collisional systems in parallel. We approach the question of algorithm scaling both from an analytic and an empirical standpoin...
详细信息
ISBN:
(纸本)1892512416
This paper looks at the performance behaviors of a methodology for simulating hard sphere collisional systems in parallel. We approach the question of algorithm scaling both from an analytic and an empirical standpoint and compare the results. In both cases, we look at results either related two or constructed from two different implementations of our methodology. The first implementation breaks the simulation up in one direction only and has been integrated into a large-scale simulation of planetary rings. The second implementation was built into a more standard granular flow model and divides the workload in two dimensions. We find that both algorithms scale nearly perfectly with the number of processors over the range that we are able to test.
In this paper, we propose an adaptive self- stabilizing algorithm for producing a d-hop connected d-hop dominating set. In the algorithm, the set is cumulatively built with communication requests between the nodes in ...
详细信息
ISBN:
(纸本)1932415262
In this paper, we propose an adaptive self- stabilizing algorithm for producing a d-hop connected d-hop dominating set. In the algorithm, the set is cumulatively built with communication requests between the nodes in the network. The set changes as the network topology changes. It contains redundancy nodes and can be used as a backbone of an ad hoc mobile network.
This paper proposes a reliability-aware mapping technique for multi applications in networks-on-chip. The proposed technique consists of three main steps: 1) Generating a new core graph enriched by spares, based on a ...
详细信息
ISBN:
(纸本)9780769549392;9781467353212
This paper proposes a reliability-aware mapping technique for multi applications in networks-on-chip. The proposed technique consists of three main steps: 1) Generating a new core graph enriched by spares, based on a given application core graph, 2) Finding smallest rectangular region to place the given application using a heuristic algorithm, and 3) Searching the specified region into whole NoC, and selecting a region which results minimum overall performance and communication energy. Spare cores are connected to all vertices of application core graph and their edges are weighted by failure probability of processing cores assigned to the application and will be updated during mapping process. Many application core graphs are used to evaluate the proposed technique. The results of 100,000 fault injection experiments show communication energy reduction and performance improvement compared to well-known related techniques in both faulty and fault-free modes.
暂无评论