Computer grids, which are formed by interconnecting computers, create vast computing resources for supporting high-end user applications in a very cost effective manner. Grid computing is an enabling technology, which is likely to underpin developments in eScience and eCommerce in the 21st century.

This unit has been updated as part of a 2009 review and comparison of data with current Handbook, Syllabus + and Callista data.

30/11/2017 - Unit disestablished at FEC 5/17 Item 7.3.

This is an elective unit for the net-centric computing major in the Bachelor of Information Technology and Systems (BITS), and for the Bachelor of Computer Science (BCompSci) and Bachelor of Software Engineering (BSE) degrees. This unit represents one of the areas of specialisation in net-centric computing. It addresses practical aspects of grid technology, notably in health sciences and business applications. For the BCompSci and BSE majors, the unit provides additional breadth in understanding distributed computing applications in a networked grid environment.

It enables students to: understand basic concepts in computer grids and clustered computing; understand some of the commonly used grid computing architectures and related performance and reliability concepts; and establish computer grids for practical use

The unit aims to produce a moderate sized pool of graduates capable of building, managing, and extending computer grids for scientific and business use.

At the completion of this unit students will be able to:

• discuss some of the enabling technologies e.g. high-speed links and storage area networks for building computer grids;
• explain the use of some of the grid computing and clustering middleware used to implement virtual super computers, including security mechanisms;
• explain programming toolkits such as Parallel Virtual Machine (PVM) and Message Passing Interface (MPI) for writing parallel computer applications;
• explain HPC Portals, peer-to-peer (P2P) networking and semantic grids;
• elaborate some of the significant grid computing areas of application e.g. Bio-Technology, eHealth and eMedicine, Finance, and Computer Networks;
• install and configure a small computer grid using Globus toolkit or a similar middleware;
• gain familiarity with commonly used grid application tools and middleware interfaces;
• extend the grid and test these applications;
• gain familiarity with MPI as employed in clusters and grids;
• understand basic performance concepts in grids and identify frequent causes of performance problems in grid applications;
• understand basic software and hardware reliability concepts in grids and identify frequent causes of reliability problems in grid applications.

020113

Contemporary computers grids provide means to implement on-demand computing. These grids can also provide a seamless (web-like) access to a variety of networked resources, e.g. large data stores and information repositories, expensive instruments, high-speed links, sensors networks, and multimedia services for a wide range of applications. Topics covered include: Computational and service-oriented grids. LVS and Beowulf Clusters. Gridservices, Webservices, WSDL, HPC Portals, Home Grids, and Peer-to-Peer (P2P) networks. Grid applications, and performance in relation to processor and network performance constraints.

On-campus

Examination (3 hours): 60%; In-semester assessment: 40%

2 hrs lectures/wk, 1 hr laboratory/wk, 1 hr tutorial/wk

Tutorial: MURPA seminar. Network lab required.

FIT

As detailed a network teaching lab with switch equipment is required for this unit to permit operation of Instant Grid software environment.

FIT1001 and one of FIT1002, CSE1202 or CSE1301 and one of FIT1005 or FIT2008

Carlo Kopp