CSC433 Parallel Systems

Cluster Computing Teaching Materials

• Cluster Computing Info Centre AND Tutorial Slides on Clusters
• Slides of Vol.1 of High Performance Cluster Computing book
• MPI Programming (Source: MHPCC)

Assignments

• Assignment 3: Cluster Programming using MPI
• Assignment 4: Study, Evaluation, and Reporting of Cluster Technologies


The New Assignment 3 and 4 Deadlines:

• Assignment 3 - 25% Due 25 May 2000.
• Assignment 4 - 25% Due 1 June 2000

Assignment 3: Solve any 3 out of the following:

• Design Parallel Sorting Algorithm and Implement using MPI; evaluate its performance for varying data size (100, 200, 1000, 5000 records) and Cluster configurations (say 1 node, 5 node, 10 node, 20 node, 40 node ...) and present results graphically.
• Design Parallel Searching Algorithm and Implement using MPI; evaluate its performance for varying data size (100, 200, 1000, 5000 records) and Cluster configurations (say 1 node, 5 node, 10 node, 20 node, 40 node ...) and present results graphically.
• Design Parallel Data Mining Algorithm for processing web logs (ask your system administrator for these files recorded for past 6 months). Identify access web access patterns, most frequently visited pages and visitors for each of the user pages (say ~/rajkumar). You may follow master slave/workers approach. Implement this algorithm using MPI; evaluate its performance for varying number of workers/slaves and Cluster configurations (say 1 node, 5 node, 10 node, 20 node, 40 node ...) and present results graphically.
• Design Parallel Matrix Multiplication Algorithm and Implement using MPI; evaluate its performance for varying data size (100x100, 200x100, 1000x2000, 5000x4000 data) and Cluster configurations (say 1 node, 5 node, 10 node, 20 node, 40 node ...) and present results graphically.
• Design Parallel Shortest Path Algorithm (say Dijkstra or Floyd Algorithm) and Implement using MPI; evaluate its performance for varying network size (10, 50, 100 nodes ) and Cluster configurations (say 1 node, 5 node, 10 node, 20 node, 40 node ...) and present results graphically.

Assignment 4: Topics for the Study, Evaluation, and Reporting of Cluster Technologies

• Cluster Operating Systems
• SolarisMC, Mosix, UnixWare, RHODOS
• Russell W Clarke <rwc@cs.monash.edu.au> and Benjamin Tee-Beng Lee <benlee@fangorn.csse.monash.edu.au>
• For report click here
• Cluster Management Systems
• Condor, PBS, Muai Scheduler, Codin, LSF, PBS. Check out: this
• Cluster Monitoring Systems
• Alert, SMILE, PARMON, bWatch.
• Cluster Tools
• Nimrod, Parallel Commands, check out Parallel Tools Consorium page AND Computing Portals.
• Colin Enticott (colin@yoyo.cc.monash.edu.au) AND Julian Crooke
• For report click here
• Cluster Programming Environments
• PVM, MPI, Linda, cJava, JavaBased
• Trevor Hansen:hansen@cs.monash.edu.au and Scott Needham:sneedham@csse.monash.edu.au
• For report click here
• Genetic Programming on Clusters
• Brian Hanly (hanly@cs.monash.edu.au) AND Daniel TUNG (beetung@cs.monash.edu.au)
• For report click here
• Cluster Applications
• Scientific, Commercial, Webserving, DataMining
• Lucas Hope <lhope@cs.monash.edu.au> and Edmund Lam (epl@cs.monash.edu.au)
• For report click here
• Clusters Systems based on Interconnection Networks
• Ehernet, Gigabit Ethernet, SCI, ATM...
• Glenn Metherall glennm@cs.monash.edu.au AND Andrew Griffiths andrewg@cs.monash.edu.au
• For report click here
• Clusters Systems based on OS
• Linux, Solaris, NT, MacOS
• Marcus Pallinger (marcusp@cs.monash.edu.au) and Richard A O Wallbrink (raw@fangorn.csse.monash.edu.au).
• For report click here
• Clusters Systems based on node architectures
• PCs, Workstations, SMPs
• Jason Ng (jng@cs.monash.edu.au) AND Greg J Rogers (gjr@cs.monash.edu.au)
• For report click here

Submission

• Students are required to submit both a hard copy and electronic submission in diskette (due 12 noon to school office).
• Also, they need to submit electronic version (source file, e.g., MS Word format) by email to me (due 5pm).

Text Books

• High Performance Cluster Computing: Architectures and  Systems,  Rajkumar Buyya (editor), Vol.1, ISBN 0-13-013784-7, Prentice Hall PTR, NJ, USA, 1999.
• Parallel Programming: Techniques and Applications Using Networked Workstations and Parallel Computers, Barry Wilkinson, C. Michael Allen, Prentice Hall, USA.
• High Performance Cluster Computing: Programming and Applications, Rajkumar Buyya (editor), Vol.2, ISBN 0-13-013785-5, Prentice Hall PTR, NJ, USA, 1999.
• Parallel Programming With Mpi, Peter Pacheco, Morgan Kaufmann Publishers, USA, 1996.
• The Review I wrote for Amazon.com on References:
  Clusters of Computers have become an appealing platform for cost-effective parallel computing and more particularly so for teaching parallel processing. At Monash University School of Computer Science and Software Engineering, I am teaching "CSC433: Parallel Systems" subject for BSc Honours students. The course covers various communication models and languages for parallel programming. Cluster Computing is one of the focused topics of this course and I found two books that suits well for this course--both published by Prentice Hall in 1999. The first one is: "High Performance Cluster Computing" by R. Buyya (editor) that I use for teaching cluster computer architecture and systems issue. The second one is "Parallel Programming" by B. Wilkinson and M. Allen that I use for teaching programming clusters using message-passing concepts. I found both books complimentary to each other.
  
  B. Wilkinson and M. Allen book discusses key aspects of parallel programming concepts and generic constructs with practical example programs. Each concept has been explained using figures and flow diagrams. The programs illustrated mostly in C using generic parallel programming constructs and popular parallel programming interfaces such as Threads, PVM, and MPI. The authors have also created an excellent web resources home page that offers presentation slides, program source codes, and instructors manual. All these tools make teaching parallel programming course, a pleasing experience. I have no hesitation in recommending this book for anyone serious about teaching parallel programming on clusters.
  
  
• Full Review on the Book, IEEE Distributed Systems Online Journal, Oct. 2001.
• Cluster Computing in the Classroom ?, IEEE/ACM International Symposium on Cluster Computing and the Grid, Brisbane, Australia, May 2001.