433-678: Cluster and Grid Computing

The knowledge of subjects such as 433-332 Operating Systems and/or 433-353 : Networks and Communications is highly prefered.  

Note: The course is open for masters(MSSE), masters by research, honours, 4th year CSSE, and research students interested in this topic. In addition, PhD students are also welcome to attend.

The growing popularity of the Internet along with the availability of powerful computers and high-speed networks as low-cost commodity components are changing the way we do parallel and distributed computing (PDC). The PDC  on local-area-networks is called  "cluster computing " and wide-area networks is called " grid computing" . Clusters employ cost-effective commodity components for building powerful computers within local-area networks, and Grids allow to share and aggregate geographically distributed resources ( such as supercomputers, storage systems, data sources, and special classes of devices) across the Internet.  They serve as cost-effective parallel and distributed computing platforms for solving large-scale resource (data and compute) intensive applications in science, commerce, and industry.

Some example of scientific and industrial applications that use these computing platforms are: system simulations, molecular modelling for drug design, brain analysis, weather forecasting, climate prediction, automobile modelling and design, structural engineering, high-energy physics, large-scale network simulation, and earth simulation.

Some examples of commercial applications that use these platforms: web servers (e.g.,  hotmail.com, yahoo.com), search engines (e.g., google.com is driven by a 4500 nodes Linux PC cluster), database engines (e.g., high-end oracle database server), financial modelling, peer-to-peer content sharing (e.g., napster) , and  web-content delivery (e.g., Akamai's world wide network of clusters delivers major web contents, e.g., cnn.com, from the nearest webserver transparently).

The IEEE/ACM Computing Curriculum 2001 has recognised the importance of net-centric computing (parallel and distributed computing) and recommended/proposed the introduction of cluster computing at senior under-graduate or post graduate level.

A coverage of topics in these two fields is discussed below:

Cluster Computing

Grid Computing

1.. Grids and Grid Technologies

2.  Programming models and Parallelization Techniques

3. Standard application development tools and paradigms such as message-passing and parameter parallel programming

4. Grid Security Infrastructure

5. Data Management

6. Application Case Study: Molecular Modelling for Drug Design and Brain Activity Analysis

7.  Resource management and scheduling

8. Setting up Grid, deployment of  Grid software and tools, and application execution.

9. Project

Project Opportunities

Assessment will be performed at three levels: 1. Study, Evaluation, and Reporting of Emerging of Cluster /Gird Technologies (10%), Project work (expected to take about 36 hours) during semester (30%), and one written examination (not exceeding 3 hours) at the end of the semester (60%). The project work must be completed satisfactorily to pass the subject. Weighting of assessment components will be advised at the commencement of the subject.

• R. Buyya (editor), High Performance Cluster Computing , Vol1. and Vol.2, Prentice Hall, USA, 1999.
  
• I. Foster and C. Kesselman (editors),  The Grid : Blueprint for a New Computing Infrastructure , Morgan Kaufmann Publishers , 1999.
  
• R. Buyya, "Economic-based Distributed Resource Management and Scheduling for Grid Computing, Ph.D. Thesis, Monash University, Melbourne, Australia, April 2002