Cluster-Handbook/Introduction

Introduction

edit

The goal of the handbook is to outline the basic concepts about computer clusters. Firstly what exactly a cluster is and how to build a simple and clean cluster which provides a great starting point for more complex tasks. Also the handbook provides a guide for task scheduling and other important cluster management features. For the practical part most of the examples use Qlustar. Qlustar is a all-in-one cluster operating system that is easy to setup, extend and most importantly operate.

A computer cluster consists of a number of computers linked together, generally through local area networks (LANs). All the connected component-computers work closely together and in many ways as a single unit. One of the big advantages of computer clusters over single computers, are that they usually improves the performance greatly, while still being cheaper than single computers of comparable speed and size. Besides that a cluster provides in general larger storage capacity, better data integrity.

Nodes

The term node usually describes one unit (linked computer) in a computer cluster. It represents the basic unit of a cluster. There a different types of notes like head-nodes, compute-, cloud- or storage-nodes. Nodes in a cluster are connected through dedicated internal Ethernet networks. The figure below shows a simple setup of the components building a basic HPC Cluster.