Tools and Environments for Parallel and Distributed Computing

Today's most complete reference on the rapidly evolving software technology used in parallel and distributed computing
Designing, developing, and utilizing the kinds of versatile computing environments that will effectively support today's complex scientific and engineering applications is an ongoing research challenge. Tools and Environments for Parallel and Distributed Computing is the first comprehensive survey of the successful computing methodologies used in today's scientific communities, including both a summary of past techniques and an introduction to future technologies.

Covering a wide variety of approaches to parallel and distributed computing, with emphasis on such factors as tradeoffs between performance and cost, the authors present a thorough survey of today's most promising software applications, their strengths and limitations, including:

• Grid computing
• Message-passing tools
• Distributed shared-memory tools
• Distributed object computing tools

Each methodology is supported by detailed case studies.
An important hands-on reference for researchers, educators, and practitioners in the field, this is a practical guide for the development and selection of today's scientific and engineering software systems

Table of Contents：

Preface.

1. Parallel and Distributed Computing (S. Hariri & M. Parashar).

1.1 Introduction: Basic Concepts.

1.2 Promises and Challenges of Parallel and Distributed Systems.

1.2.1 Processing Technology.

1.2.2 Networking Technology.

1.2.3 Software Tools and Environments.

1.3 Distributed System Design Framework.

References and Further Reading.

2. Message-Passing Tools (S. Hariri & I. Ra).

2.1 Introduction.

2.2 Message-Passing Tools versus Distributed Shared Memory.

2.2.1 Distributed Shared Memory Model.

2.2.2 Message-Passing Model.

2.3 Message-Passing System: Desirable Features.

2.4 Classification of Message-Passing Tools.

2.4.1 Classification by Implementation.

2.5 Overview of Message-Passing Tools.

2.5.1 Socket-Based Message Passing.

2.5.2 p4.

2.5.3 Parallel Virtual Machine.

2.5.4 Message-Passing Interface.

2.5.5 Nexus.

2.5.6 Madeleine I and II.

2.5.7 Active Messages.

2.6 ACS.

2.6.1 Multithread Communications Services.

2.6.2 Separation of Data and Control Functions.

2.6.3 Programmable Communication, Control, and Management Service.

2.6.4 Multiple Communication Interfaces.

2.6.5 Adaptive Group Communication Services.

2.7 Experimental Results and Analysis.

2.7.1 Experimental Environment.

2.7.2 Performance of Primitives.

2.7.3 Application Performance Benchmarking.

2.7.4 Performance Results of Adaptive Schemes.

2.8 Conclusions.

References.

3. Distributed Shared Memory Tools (M. Parashar & S. Chandra).

3.1 Introduction.

3.2 Cache Coherence.

3.2.1 Directory-Based Cache Coherence.

3.3 Shared Memory Consistency Models.

3.4 Distributed Memory Architectures.

3.5 Classification of Distributed Shared Memory Systems.

3.5.1 Hardware-Based DSM Systems.

3.5.2 Mostly Software Page-Based DSM Systems.

3.5.3 All-Software Object-Based DSM Systems.

References.

4. Distributed-Object Computing Tools (R. Raje, et al.).

4.1 Introduction.

4.2 Basic Model.

4.2.1 RMI.

4.2.2 CORBA.

4.2.3 DCOM.

4.3 Examples.

4.3.1 Experimental Setup.

4.3.2 Developing Applications under RMI, CORBA, and DCOM.

4.3.3 Experiment 1: Ping.

4.3.4 Experiment 2: Producer-Consumer Problem.

4.3.5 Experiment 3: Numerical Computation.

4.4 Comparison of the Three Paradigms.

4.4.1 Dependency Issues.

4.4.2 Implementation Details.

4.4.3 Architecture Details.

4.4.4 Support for Additional Features.

4.4.5 Performance Comparison.

4.5 Conclusions.

References.

5. Gestalt of the Grid (G. von Laszewski & P. Wagstrom).

5.1 Introduction.

5.1.1 Motivation.

5.1.2 Enabling Factors.

5.2 Definitions.

5.3 Multifaceted Grid Architecture.

5.3.1 N-Tiered Grid Architecture.

5.3.2 Role-Based Grid Architecture.

5.3.3 Service-Based Grid Architecture.

5.3.4 Grid Challenges.

5.4 Grid Management Aspects.

5.4.1 Managing Grid Security.

5.4.2 Managing Grid Information.

5.4.3 Managing Grid Data.

5.4.4 Managing Grid Execution and Resources.

5.4.5 Managing Grid Software.

5.4.6 Managing Grid Hardware.

5.5 Grid Activities.

5.5.1 Community Activities.

5.5.2 Grid Middleware.

5.5.3 High-Throughput Computing.

5.6 Grid Applications.

5.6.1 Astrophysics Simulation Collaboratory.

5.6.2 Particle Physics Data Grid.

5.6.3 NEESgrid.

5.7 Portals.

5.7.1 HotPage.

5.7.2 Webflow and Gateway.

5.7.3 XCAT.

5.7.4 UNICORE.

5.7.5 JiPANG.

5.7.6 PUNCH.

5.7.7 Access Grid.

5.7.8 Commercial Grid Activities.

5.8 Conclusions.

References.

6. Software Development for Parallel and Distributed Computing (M. Parashar & S. Hariri).

6.1 Introduction.

6.2 Issues in HPC Software Development.

6.2.1 Models for Parallel Computation.

6.2.2 Portable Application Description Medium.

6.2.3 Parallel Algorithm Development.

6.2.4 Program Implementation and Runtime.

6.2.5 Visualization and Animation.

6.2.6 Maintainability.

6.2.7 Reliability.

6.2.8 Reusability.

6.3 HPC Software Development Process.

6.4 Parallel Modeling of Stock Option Pricing.

6.5 Inputs.

6.6 Application Analysis Stage.

6.7 Application Development Stage.

6.7.1 Algorithm Development Module.

6.7.2 System-Level Mapping Module.

6.7.3 Machine-Level Mapping Module.

6.7.4 Implementation/Coding Module.

6.7.5 Design Evaluator Module.

6.8 Compile-Time and Runtime Stage.

6.9 Evaluation Stage.

6.10 Maintenance/Evolution Stage.

6.11 Existing Software Support.

6.11.1 Application Specifications Filter.

6.11.2 Application Analysis Stage.

6.11.3 Application Development Stage.

6.11.4 Compile-Time and Runtime Stage.

6.11.5 Evaluation Stage.

6.11.6 Maintenance/Evolution Stage.

References. 

Index

今天最完整的參考資料，關於在並行和分散式計算中快速發展的軟體技術。設計、開發和利用多功能計算環境，以有效支援當今複雜的科學和工程應用，是一個持續的研究挑戰。《並行和分散式計算的工具和環境》是對當今科學社群中成功使用的計算方法的首次全面調查，包括對過去技術的總結和對未來技術的介紹。

本書涵蓋了各種並行和分散式計算方法，強調性能和成本之間的權衡，作者們對當今最有前景的軟體應用進行了全面調查，包括它們的優點和限制，其中包括：

- 網格計算
- 消息傳遞工具
- 分散式共享內存工具
- 分散式物件計算工具

每種方法都有詳細的案例研究支持。這是一本對該領域的研究人員、教育工作者和從業人員來說非常重要的實用參考資料，也是選擇和開發當今科學和工程軟體系統的實用指南。

目錄：

前言

第1章 並行和分散式計算（S. Hariri和M. Parashar）
1.1 簡介：基本概念
1.2 並行和分散式系統的優勢和挑戰
1.2.1 處理技術
1.2.2 網絡技術
1.2.3 軟體工具和環境
1.3 分散式系統設計框架
參考文獻

第2章 消息傳遞工具（S. Hariri和I. Ra）
2.1 簡介
2.2 消息傳遞工具與分散式共享內存的比較
2.2.1 分散式共享內存模型
2.2.2 消息傳遞模型
2.3 消息傳遞系統：理想特性
2.4 消息傳遞工具的分類
2.4.1 根據實現方式的分類
2.5 消息傳遞工具概述
2.5.1 基於套接字的消息傳遞
2.5.2 p4
2.5.3 平行虛擬機器
2.5.4 消息傳遞接口
2.5.5 Nexus
2.5.6 Madeleine I和II
2.5.7 主動消息
2.6 ACS
2.6.1 多線程通信服務
2.6.2 數據和控制功能的分離
2.6.3 可編程通信、控制和管理服務
2.6.4 多種通信接口
2.6.5 自適應群組通信服務
2.7 實驗結果和分析
2.7.1 實驗環境
2.7.2 基本操作的性能
2.7.3 應用性能基準測試
2.7.4 自適應方案的性能結果
2.8 結論
參考文獻

第3章 分散式共享內存工具（M. Parashar和S. Chandra）
3.1 簡介
3.2 快取一致性
3.2.1 基於目錄的快取一致性
3.3 共享內存一致性模型
3.4 分散式內存架構
3.5 分散式共享內存系統的分類
3.5.1 基於硬體的DSM系統
3.5.2 大部分軟體基於頁的DSM系統
3.5.3 全軟體基於物件的DSM系統
參考文獻

第4章 分散式物件計算工具（R. Raje等）
4.1 簡介
4.2 基本模型
4.2.1 RMI
4.2.2 CORBA
4.2.3 DCOM
4.3 實例
4.3.1 實驗設置
4.3.2 在RMI、CORBA和DCOM下開發應用程式
4.3.3 實驗1：Ping
4.3.4 實驗2：生產者-消費者問題
4.3.5 實驗3：數值計算
4.4 三種範式的比較
4.4.1 依賴性問題
4.4.2 實現方式比較