Students on campus

MUGrid Bootcamp

Receive training for Marquette University's research computing resources available on MUGrid. The bootcamp will be held July 14th and 15th, 2010. Participants can attend any number of the sessions described below.

Location:
Cudahy Hall, Room 412
1313 West Wisconsin Avenue
on the campus of Marquette University

Schedule:

Wednesday, July 14th, 2010
8:30-9:00Registration
9:00-10:15Session 1: Accessing MUGrid and Command Line Basics (Notes, Presentation)
10:15-10:30Break
10:30-11:45Session 2: Shell Scripting (Notes, Presentation)
11:45-1:30Lunch (on your own)
1:30-2:45Session 3: Basic High-Throughput Computing (Presentation)
2:45-Individual Help
Thursday, July 15th, 2010
8:30-9:00Registration
9:00-10:15Session 4: Parallel Computing on MUGrid (Presentation)
10:15-10:30Break
10:30-11:45Session 5: MATLAB on MUGrid (Presentation)
11:45-1:30Lunch (on your own)
1:30-2:45Session 6: Advanced High-Throughput Computing (Presentation)
2:45-3:00Break
3:00-4:00Session 7: Using National Grids (Presentation)

Session Descriptions:

Session 1: Access to MUGrid and Command Line Basics Instructions for accessing MUGrid resources using secure shell clients for various platforms (Windows, Mac OS X, Linux). Includes information on how to transfer files to and from MUGrid resources. Covers the basics of using the Linux command, the layout of the Linux file system, essential commands, etc., which are the primary tools for interacting with MUGrid.

Session 2: Shell Scripting Introduction to shell scripting, particularly with the bash shell. Includes basic control structures (loops and conditions), practice generating text files, simple calculations, etc. Requires knowledge of a Linux text editor and familiarity with the Linux command line, which are covered in Session 1.

Session 3: Basic High-Throughput Computing Introduction to high-throughput computing using the Condor workload management system. Covers how to design, submit, and manage simple jobs, including basic parameter sweeps, simple Monte Carlo simulations, etc. Requires knowledge of material covered in Sessions 1 and 2.

Session 4: Parallel Computing on MUGrid How to implement and run parallel applications on MUGrid using the message passing interface (MPI), Condor, and the PBS scheduling system. Requires knowledge of material covered in Sessions 1 and 2. Session 3 is recommended for users interested in mixing parallel and high-throughput computing.

Session 5: MATLAB on MUGrid How to prepare MATLAB applications for execution on MUGrid. Includes compiling MATLAB vode, passing parameters through Condor, using external MATLAB toolkits, etc. Requires experience with MATLAB and knowledge of the materials from Sessions 1-3.

Session 6: Advanced High-Throughput Computing Automating and managing scientific workflows with Condor's DAGMAN tool. Covers issues related to data staging, job dependencies, workflow recovery, robustness, etc. Requires knowledge of materials from Sessions 1-3.

Session 7: Using National Grids An introduction to TeraGrid, Open Science Grid, and other national grids. Covers the kinds of resources available on national resources and how to gain access to those resources. Knowledge of material in Session 1 will be helpful, but not required

Individual Help Participants are encouraged to bring their own research projects with them to receive individualized help in planning and preparation for using MUGrid. Help may take the form of discussion, ad-hoc coding, and scheduling appointments for further meetings.

Please respond by July 12 to Craig Struble at cstruble-bootcamp@mscs.mu.edu

Parking is available in the 16th Street Parking Structure located between West Wisconsin Avenue and West Wells Street and the Wells Street Parking Structure located between 12th and 13th streets.

Questions or special needs, please contact Craig Struble as described above.


SITE MENU

Circuit Board

Campus Grids

Campus grids link together computing resources at an institution to support research and collaboration. A goal of campus grids is to provide a seamless workflow from data collection to analysis and dissemination of results. The campus grid is an essential component of discovery in the 21st Century. Read the Cyberinfrastructure Vision for 21st Century Discovery report for more details.