Clustering - Part 1

Table of contents:

1. 1 Introduction
2.     1.1 Goals
3. 2 Clusters
4.     2.1 Different type of Clusters
5.     2.2 Different type of Nodes
6.     2.3 Clusters vs SMP/MPP
7.     2.4 Grid
8. 3 Software
9.     3.1 Operating systems
10.     3.2 Management Systems
11.     3.3 Databases
12. Bibliography
13.     References
14.     Recommended articles

Summary
   This article is aiming to present the Clusters and the different methods developed to generate massively parallel computing solutions. We will show that the hardware and associated software solutions depend on the desired goals.
We will describe 3 types of Clusters:

1. High Performance Computing
2. High Viability
3. High Scalarity

    Then, we will have a look at the different node systems available and their respective functions. We will compare Massively Parallel Processor solutions (MPP) to Grid computing.
    We will not neglect the software part. We will first look at MPI, PVM and AFAPI, as well as a management system for parallel computing and load splitting system, OpenMosix. We will then quickly describe the concept of resource virtualization for a cluster that can be consider as a management system by itself.
   Finally we will try to explain the different operating systems which can be buzzed for clustering, their respective strength and weakness.