Small failures should only disrupt a small part of a network. One way to do this is by marking the surrounding area as untrustworthy - circumscribing the failure. This can be done with a distributed algorithm using hi...
详细信息
ISBN:
(纸本)088986392X
Small failures should only disrupt a small part of a network. One way to do this is by marking the surrounding area as untrustworthy - circumscribing the failure. This can be done with a distributed algorithm using hierarchical clustering and neighbor relations, and the resulting circumscription is near-optimal for convex failures.
The proceedings contains 85 papers from the Fifteenth iasted international conference on parallel and distributed computing and systems: Volume I. The topics discussed include: divisible load scheduling for grid compu...
详细信息
ISBN:
(纸本)088986392X
The proceedings contains 85 papers from the Fifteenth iasted international conference on parallel and distributed computing and systems: Volume I. The topics discussed include: divisible load scheduling for grid computing;comparison of centralized and decentralized service discovery in a grid environment;reliable multicast protocols for Java-based grid middleware platforms;self-correcting broadcast in distributed hash tables;adaptive file-grain distribution in a cluster file system;and a request-based self-stabilizing token passing.
Global predicate detection is a fundamental problem in distributedcomputing in the areas of distributed debugging and software fault-tolerance. It requires searching the global state lattice of a computation to deter...
详细信息
ISBN:
(纸本)088986392X
Global predicate detection is a fundamental problem in distributedcomputing in the areas of distributed debugging and software fault-tolerance. It requires searching the global state lattice of a computation to determine if any consistent global state satisfies the given predicate. We give an efficient algorithm that perform the lex traversal of the lattice. We also give a space efficient algorithm for the breadth-first-search (BFS) traversal.
Today, Grid is one of major aspects in distributedcomputing environment, whose driving force is attributed to high performance computing, or CPU-cycle consuming computation. As a programming environment on Grid, we i...
详细信息
ISBN:
(纸本)088986392X
Today, Grid is one of major aspects in distributedcomputing environment, whose driving force is attributed to high performance computing, or CPU-cycle consuming computation. As a programming environment on Grid, we introduce Hem(π), and propose using MPI as a programming model on a supercomputing network. This MPI programming environment works on a heterogeneous distributed environment, and can represent large scale resource consuming computation. Shape is used to define computing in Hem(π). In Hem(π), shape define computation as a coupled system of local MPI tasks and servers. Furthermore, remote accesses to memory and files which play central role in distributedcomputing are discussed.
Adaptive task partitioning may introduce implicit hot spots in the computations which affect the parallel performance of the adaptive multivariate integration algorithm. To maintain parallel efficiency, one needs to b...
详细信息
ISBN:
(纸本)088986392X
Adaptive task partitioning may introduce implicit hot spots in the computations which affect the parallel performance of the adaptive multivariate integration algorithm. To maintain parallel efficiency, one needs to balance the load of the processors by distributing the computationally intensive parts of the problem domain. We study and improve the performance of the scheduler based or centralized load balancing method in the adaptive task partitioning algorithm implemented in ParInt.
In this paper, we propose an algorithm to solve the group k-mutual exclusion (Gk-ME) problem for distributedsystems. The Gk-ME problem is concerned with controlling the concurrent accesses of some resource by at most...
详细信息
ISBN:
(纸本)088986392X
In this paper, we propose an algorithm to solve the group k-mutual exclusion (Gk-ME) problem for distributedsystems. The Gk-ME problem is concerned with controlling the concurrent accesses of some resource by at most k nodes with the constraint that no two distinct resources can be accessed simultaneously. The proposed algorithm utilizes a token circulation mechanism and does not require the nodes to have identifications. The delay and session switches of the proposed algorithm are Ω(n2) and Ω(n), respectively.
The use of divisible load scheduling theory to model and design grid systems such as those arising in large physics experiments is discussed. Current divisible load theory is summarized. A typical application, the STA...
详细信息
ISBN:
(纸本)088986392X
The use of divisible load scheduling theory to model and design grid systems such as those arising in large physics experiments is discussed. Current divisible load theory is summarized. A typical application, the STAR experiment at the Relativistic Heavy Ion Collider (RHIC) is discussed. This includes a sample calculation based on existing infrastructure numbers.
In this paper, we present a new auction algorithm for the linear assignment problem, based on a new technique, called look-back bidding. The main idea is to keep local working sets of objects. It leads to significant ...
详细信息
ISBN:
(纸本)088986392X
In this paper, we present a new auction algorithm for the linear assignment problem, based on a new technique, called look-back bidding. The main idea is to keep local working sets of objects. It leads to significant performance gains both in sequential and distributed memory parallel versions. We give performance results of the new algorithm, called look-back auction algorithm, and compare its complexity with standard auction algorithms. In the parallel version, our new algorithm eliminates a substantial amount of interprocessor communication.
We consider the problem of determining parallel complexity of solving banded triangular linear systems using substitution on a k-dimensional torus network. We present lower bounds on execution time for solving these s...
详细信息
ISBN:
(纸本)088986392X
We consider the problem of determining parallel complexity of solving banded triangular linear systems using substitution on a k-dimensional torus network. We present lower bounds on execution time for solving these systems, taking into account communication costs. Furthermore, optimal algorithms are designed.
We introduce a new view of distributed computation, called the NavP view, under which a distributed program is composed of multiple sequential self-migrating threads called DSCs. In contrast with those in the conventi...
详细信息
ISBN:
(纸本)088986392X
We introduce a new view of distributed computation, called the NavP view, under which a distributed program is composed of multiple sequential self-migrating threads called DSCs. In contrast with those in the conventional SPMD style, programs developed in the NavP view exhibit the nice properties of algorithmic integrity and parallel program composition orthogonality, which make them clean and easy to develop and maintain. The NavP programs are also scalable. We use example code and performance data to demonstrate the advantages of using the NavP view for general purpose distributedparallel programming.
暂无评论