This report discusses the current state of the NYSGRID. Which institutions are on the NYSGRID along with their status. How do we get new institutions on the grid, repair those which are broken and upgrade those who are running old software stacks.

close

Grid technology provided by the Center for Computation Reseach (CCR) through NYSGrid organization (http://www.nysgrid.org) is supporting aerospace engineering research to Professor Vadrevu Murthy from the department of Mechanical and Aerospace Engineering at the L.C. Smith College of Engineering and Computer Science at Syracuse University. The rotary-wing unsteady aerodynamic group lead by Prof. Murthy uses grid resources to perform numerical simulations of three-dimensional dynamic stall to advance helicopter rotor technology. Ercan Dumlupinar will present current research going on at Syracuse University, and the need for computationally intensive jobs at aerospace field.

close

Porting commercial applications to run over grid computing infrastructure can be a challenge. This is especially true for licensed applications and those requiring MPI. We will present challenges and how they were addressed using two case studies: Cobalt Computational Fluid Dynamic software and and Autodock.

close

Creating software models of the human anatomy and imaging systems, and modeling the medical physics of the imaging acquisition process can provide a means to generate realistic synthetic data sets. In many cases synthetic data sets can be used, reducing the time and cost of collecting real images, and making data sets available to institutions without clinical imaging systems.

While medical image simulation software has been under development since the 1980s, until recently the complexity of the procedures and long computation times have limited the realism and accuracy of artificially generated images. The use of distributed systems such as the Blue Gene/L systems available at Brookhaven National Labs and the Rensselaer Polytechnic Institute provide unparalleled computational capabilities allowing full resolution data sets to be generated in reasonable amounts of time.

SIMRI is a Bloch equation based magnetic resonance image (MRI) simulator, developed at CREATIS, Lyons, France, designed for small clusters supporting MPI. Modifications were implemented to support the memory constraints and increased number of nodes available on the Blue Gene/L Systems. This resulted in, to the author's knowledge, the first system capable of generating high resolution (2563 voxels) realistic MRI images directly from modeling the Bloch equations.

close

Upgrading to OSG 1.0.0 is important because it contains several new features and bug-fixes from previous versions.

close

Several examples will be provided to give attendees experience with various concepts needed to run applications on the grid. These include the staging of data, polling for status, retrieving results, and cleaning up. This workshop will give attendees example scripts and hands-on experience running grid applications.

Note: In order to participate in the hands-on portion of the workshop, attendees must have previously obtained a DoE certificate and must be a member of the NYSGRID VO. Attendees will also need access to a server with the OSG tools installed in order to submit the example applications. Please see the instructions for obtaining a certificate and joining the NYSGRID VO at http://nysgrid.org/main/userinfo/index.maml

close

Leverage customers existing knowledge of condor to submit and track grid jobs. This discussion will cover the admin side of the equation. How to setup condor ad injection and how it simplifies the grid scheduling problem by giving users a homogeneous interface to the heterogeneous pool of grid resource. Discussion will also be seeking a way for NYSGrid sites to tackle the problem of advertizing up-to-date statistics for their grid site, push or pull statistics gathering, and other issues.

close

Taking computational python scripts and scaling to the size of the grid. Targeting end users as well as systems administrators, I will demonstrate recipes that can be used to scale a python script from one standalone system to the grid. If there is time I will also demonstrate that the same system works on campus clusters of windows systems running condor.

close

In this presentation, we describe a service oriented architecture and Grid abstraction framework that allows us to access Grids through JavaScript. Obviously, such a framework integrates well with other Web 2.0 technologies. The framework consists of two parts. A client Application Programming Interface (API) to access the Grid via JavaScript and a mediator service and API through which the Grid access is channeled.

close

The Laser Interferometer Gravitational-wave Observatory (LIGO) is an ambitions NSF-funded project to detect gravitational waves and to use these wave to study the Universe. Syracuse University is a member of the LIGO Scientific Collaboration (LSC) and is contributing to the search for gravitational waves in LIGO data. This requires the transfer of multi-terabyte data sets to the Syracuse University Gravitation and Relativity (SUGAR) cluster, a 320 CPU core supercomputer in the SU Physics Department. Syracuse, NYSERNet and the LSC group at UW-Milwaukee have collaborated to connect the Syracuse and Milwaukee clusters over the Internet2 Dynamic Circuit Network (DCN) to transfer LIGO data to the SUGAR cluster. We describe the LIGO data transfer problem at Syracuse and how we have successfully used DCN to solve it.

close