|Long Title:||Condor Desktop Grid Integration into a National Computing Grid|
|Project Leader:||M. Thiémard|
|Deputy Project Leader:||P. Jermini|
The product developed in this project enables sites to interface Desktop Grids (also called Campus
Grids) running the Condor middleware to the Swiss Multi-Science Computing Grid (SMSCG).
Users may download the documentation "Setting Up Condor on ARC infrastructure" and use the ARC patches for Condor (available through SMSCG repository, restricted access).
By giving access to Campus Grids via SMSCG, we ensure a diversity in the types of resources available to researchers: most of the resources available in SMSCG are typical high-performance cluster, tailored for parallel applications, whereas Campus Grids focus more on sequential applications, where there is no need of communications between the different computing hosts.
In order to ease the collaboration with the SMSCG project, EPFL has joined the project SMSCG II, with the intent to provide computing resources. This task will also be made easier by the set up of a Condor sub-pool entirely managed by Central IT Services (unlike the current configuration).
The goal of this project was to integrate a local Desktop Grid into a wider national grid effort (such as the Swiss Multi-Science Computing Grid). Indeed, there was at that moment no such type of resources integrated into a swiss national grid. This project allows a diversification of available resources, and to promote this kind of resources for other institutions wishing to join their local desktop grids into the wider Swiss Multi-Science Computing Grid (SMSCG).
Main step has been the ARC-Condor integration, providing an interoperability platform between Condor (middleware used at EPFL) and ARC (common middleware, used by SMSCG), in line with the other interoperability guidelines defined by the SMSCG project. This platform should allow seamless execution of jobs coming from the grid infrastructure on Condor pools. Also a new scientific field was brought to the SMSCG project, namely cryptography, where the needs of computational power are very high.