the generic functionality of passive testing is detecting failures (fault manifestations) in a distributed system. We deal with one of the key functions of a passive tester - Trace Analysis (TA). the standard TA algor...
详细信息
ISBN:
(纸本)0889865701
the generic functionality of passive testing is detecting failures (fault manifestations) in a distributed system. We deal with one of the key functions of a passive tester - Trace Analysis (TA). the standard TA algorithms proceed by traversing a Reference Specification (RS), which is usually given in the form of a FSM automaton. the run-time performance of such algorithms is known to be poor for a large, nondeterministic RS (i.e. for real protocols). To improve this aspect of passive testing, instead of considering new TA algorithms, we try to increase the performance of existing algorithms, by changing the form of a RS to an Event Graph (EG). We define an EG and submit a detailed algorithm for its derivation. We then outline the EG-based algorithm for RS traversal. We report on the results of simulation experiments that clearly show considerable improvements in run-time performance of EG-based passive testing over previous approaches, while the functionality of failure detection is strictly retained.
this paper presents coordinated virtual partition (CVP) for Grid computingsystems. the CVP is a way for regulating the resources supplied to different components of an application in unison according to an agreed rel...
详细信息
this paper presents coordinated virtual partition (CVP) for Grid computingsystems. the CVP is a way for regulating the resources supplied to different components of an application in unison according to an agreed relative proportion. this study shows that coordinated resource provisioning has several benefits including: (a) reducing the wait times experienced by an application and (b) improving the overall application performance by reducing the wait times. the CVP achieves these benefits by releasing resources from "fast" running application components that can be reallocated by the Grid for other applications.
Due to the prevalence of sensors such as live cameras or environmental sensors, sensor data stream delivery, which requires continuous and cyclic data delivery attracts great attention. For sensor data stream delivery...
详细信息
ISBN:
(纸本)9781467345651;9780769549033
Due to the prevalence of sensors such as live cameras or environmental sensors, sensor data stream delivery, which requires continuous and cyclic data delivery attracts great attention. For sensor data stream delivery, various communication loads balancing techniques have been studied since the load of the sensor data source become high to accommodate large number of clients. However, these studies assume only the requests that have the same collection cycle, which is not enough for the actual applications. In this paper, we propose a sensor data stream delivery system with communication loads balancing for heterogeneous collection cycle requests. the proposed system distributes the loads by re-delivering the sensor data that are requested by other clients with different collection cycles but have common cycles.
the SPACE RIP technique is one of the parallel imaging methods that has the potential to revolutionize the field of fast MR imaging. the image reconstruction problem of SPACE RIP is a computation intensive task which ...
详细信息
the SPACE RIP technique is one of the parallel imaging methods that has the potential to revolutionize the field of fast MR imaging. the image reconstruction problem of SPACE RIP is a computation intensive task which needs to be parallelized to further reduce the reconstruction time. In this paper, we analyzed the algorithm and identified the program bottleneck to be parallelized. the loop level parallelization is implemented with Pthread, OpenMP and MPI. Furthermore, since the reconstruction uses Singular Value decomposition (SVD) to solve the matrix pseudoinverse problem, we implemented the one sided Jacobi parallel SVD on the state-of-art cellular computer architecture Cyclops64 to speedup the problem at the fine grain level.
this paper presents a new generalized cellular automata (GCA) approach to effectively solve a class of optimization problems subject to a binary constraint matrix1. In contrast to the Hopfield-type neural network (HNN...
详细信息
Resource management is an important aspect in open distributedsystems, as these systems are persistent and ubiquitous. In order to be scalable, it is imperative that the resource management itself should be distribut...
详细信息
Resource management is an important aspect in open distributedsystems, as these systems are persistent and ubiquitous. In order to be scalable, it is imperative that the resource management itself should be distributed as well as the resources. this paper focuses on the LINDA coordination model of open distributedsystems. One limited resource is memory, and garbage collection has already been proposed for the standard LINDA with multiple tuple-spaces (TSs) to avoid memory exhaustion. the implementation, however, was restricted to garbage collection of TSs. Taking into account the need for garbage collection not only for TSs, but also for tuples, this paper demonstrates how this can be extended to tuples, withthe introduction of multicapabilities, which generalise capabilities to collections of objects. We also illustrate the use of multicapabilities in two other applications related to resource management: managing deadlocks and information caching.
the balance between CPU speed and interconnection network throughput in distributed memory parallel computers varies with each generation of systems, but the trend is that CPUs are gaining performance faster than the ...
详细信息
the balance between CPU speed and interconnection network throughput in distributed memory parallel computers varies with each generation of systems, but the trend is that CPUs are gaining performance faster than the interconnection networks. this means that remote data accesses are becoming more expensive relative to local accesses in terms of CPU cycles. therefore, remote memory access mechanisms that were suited to a previous generation of parallel machines may be less appropriate for current clusters. this research evaluates a multithreaded programming paradigm with cached remote memory accesses and thread migration to exploit array locality on a cluster with Myrinet. the approach, called Nomadic threads, was originally developed for the CM5, but has been adapted to use MPI on Linux clusters. the results show that the current surfeit of CPU power vs. network throughput dramatically changes scaling characteristics of some programs while others behave much as they did on the decade-old CMS.
Multi-cluster schedulers can dramatically improve average job turn-around time performance by making use of fragmented node resources available throughout the grid. By carefully mapping jobs across potentially many cl...
详细信息
ISBN:
(纸本)9780889867048
Multi-cluster schedulers can dramatically improve average job turn-around time performance by making use of fragmented node resources available throughout the grid. By carefully mapping jobs across potentially many clusters, jobs that would otherwise wait in the queue for local cluster resources can begin execution much earlier;thereby improving system utilization and reducing average queue waiting time. Recent research in this area leverages user-provided estimates of job communication characteristics to effectively partition the job across cluster boundaries. In this paper, we address the impact of inaccuracies in these estimates on overall system performance. Furthermore, we demonstrate that multi-site job scheduling techniques benefit from these estimates, even in the presence of considerable inaccuracy.
the proceedings contain 40 papers. the topics discussed include: hybrid algorithms for list ranking and graph connected components;parallel multiple precision division by a single precision divisor;scalable clustering...
ISBN:
(纸本)9781457719516
the proceedings contain 40 papers. the topics discussed include: hybrid algorithms for list ranking and graph connected components;parallel multiple precision division by a single precision divisor;scalable clustering using multiple GPUs;hybrid implementation of error diffusion dithering;porting irregular reductions on heterogeneous CPU-GPU configurations;building algorithmically nonstop fault tolerant MPI programs;high-level template for the task-based parallel wavefront pattern;coordination mechanisms for selfish multi-organization scheduling;maximizing throughput of jobs with multiple resource requirements;scheduling diverse high performance computingsystems withthe goal of maximizing utilization;a dynamic scheduling framework for emerging heterogeneous systems;improving graph coloring on distributed-memory parallel computers;and multi-model prediction for enhancing content locality in elastic server infrastructures.
In this paper, we study a parallel job scheduling model which takes into account both computation time and the overhead from communication between processors. Assuming that a job Jj has a processing requirement pj and...
详细信息
ISBN:
(纸本)9780889867048
In this paper, we study a parallel job scheduling model which takes into account both computation time and the overhead from communication between processors. Assuming that a job Jj has a processing requirement pj and is assigned to kj processors for parallel execution, then the execution time will be modeled by tj = p j / kj+ (kj - 1) c, where c is the constant overhead cost associated with each processor other than the master processor. In this model, (kj - 1)c represents the cost for communication and coordination among the processors. this model attempts to accurately portray the actual execution time for jobs running in parallel on multiple processors. Using this model, we will study the online algorithm Earliest Completion Time (ECT) and show a lower bound for the competitive ratio of ECT for m ≥ 2 processors. For m ≤ 4, we show the matching upper bound to complete the competitive analysis for m = 2,3,4. For large m, we conjecture that the ratio approaches 30/13 ≈ 2.30769.
暂无评论