Cluster Concepts

From Cluster Documentation Project
Revision as of 11:43, 12 May 2014 by Deadline (talk | contribs) (fixed CM links)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Building Machines Around Problems[edit]

Cluster computing is a exciting technology. It allows users to design a computing resource around a problem. In the past problems were often designed around large monolithic supercomputers. A typical supercomputer has a large amount of costly engineering and design that may or may not be needed for a specific application. Because the supercomputer comes in a "big box", the end user must pay for the cost of all the technology. Clusters built from commodity components allow the user to only buy what they needed at commodity prices. The use of clustered computing has redefined the price-to-performance curve for HPC (High Performance Computing).

A good working definition of a cluster was first given in the book How to Build a Beowulf: A Guide to the Implementation and Application of PC Clusters by Thomas Sterling, John Salmon, Donald J. Becker and Daniel F. Savarese

A Beowulf is a collection of personal computers (PCs) interconnected by widely available networking technology running one of several open-source Unix like operating systems.

Cluster Monkey has a list of cluster books.

If you are new to the cluster community, you may find the Getting Started with Clusters series on Cluster Monkey helpful.