Building Linux Clusters

Just as the workstation has replaced the mainframe, today's PCs handle an increasing portion of the world's server load. Nowhere is this more evident than in the phenomenal growth of the Linux operating system. Today Linux reaches into the high performance realm of parallel computing by means of clustering.

Building Linux Clusters is a primer for the veteran Linux programmer who wants to take those first steps into parallel computing. From graphics rendering farms to fault tolerant database servers, Linux clusters can extend Linux solutions into high performance and high availability computing.

If you're an experienced C programmer with little exposure to parallel computing, but want to learn the basics of cluster installation and cluster management, Building Linux Clusters is essential reading. This book describes how to install basic Red Hat Linux modified for clustering, how to proliferate installation from master to slave nodes, and how to set up cluster management tools.

You'll also learn how to manage a clustered system. You'll get an introduction to parallel programming strategies, helping you to identify which problems are right for a clustered solution. In addition, the book covers the Open Source tools available for parallel development and parallelizing existing applications.

Inside you'll also find a CD, which includes:

• Red Hat 6.2
• Linux clustering software
• Cluster management tools and scripts
• Parallel programming tools

With thirteen years' work in network design and security systems, David HM Spector brings a wealth of expertise to the area of high performance clustered computing. In Building Linux Clusters, he provides a complete set of tools to introduce you to clustered solutions.

1. Cluster Design, Development, and Management
1. Introduction
• What Makes a Supercomputer "Super?"
• Pushing the State of the Art
• The Trickle-Down Effect
• Science Fiction Becomes Science Fact
• What Can Clusters Be Used For?
• What's Next?
2. Basic Concepts
• Why Clusters?
• Clustering Concepts
• Networking Concepts
• TCP/IP Addressing
• Parallel Programming Systems
• Review
3. Designing Clusters
• Design Considerations
• Hardware for Clusters
• System Performance Analysis
• Disk Selection
• Network Selection
• Cluster Configurations
• Scalability
• Data Access for Clusters
• Messaging Systems
• Queueing Systems
• Compilers
• Other Tools
4. Building Clusters
• Selecting the Place
• Preparing the Environment
• Building the Nodes
• Assembling Custom Systems
• Installation and Cabling
• Summary
5. Software Installation and Configuration
• CD-ROM Overview
• Installation Overview
6. Managing Clusters
• Basic Tools
• The Cluster Management System
• Structure of the Management System
• The Cluster Management Page
• User Administration
• Group Administration
• Project Administration
• Batch Queues
• Accounting
• Cluster Maintenance
• Other Management Interfaces
• Other Useful Tools
• Summary
2. Cluster Programming and Applications
7. Tools and Libraries for Parallel Programming
• Development Tools
• Parallel Development Environments
• Parallel Libraries
• Debugging/Profiling Tools
• Online Documentation
• System Extensions
• Summary
8. Programming in a Parallel Environment
• Programming a Clustered System
• A Clustered MUD
• Language Selection
• Where to Go from Here
9. Application Examples
• mp3pvm
• PVMPOV
• PVFS
3. Appendixes
1. Resources
2. Message Passing APIs
3. Installation Scripts
4. The Cluster Administration Database

If you really want this book, you need to get the re-written chapter 2 from the O'Reilly's website, as the original is totally messed up (in fact, so messed up that O'Reilly has withdrawn the book). Otherwise, it is a decent book on how to build so-called BeoWulf class machines.