SCI: Keeping Condor Flight Worthy

点击翻译按钮获取中文摘要

This proposal outlines an effort to sustain the Condor High Throughput Computing Software (Condor) by the Condor Project at the University of Wisconsin-Madison (UW). The activities proposed are categorized into four areas of work: Support, Release, Enhance, and Use. "Support" activities aid users with installation, configuration, and usage through ticket-tracked email. Support also includes outreach work in the scientific computing community. This outreach work includes hosting an annual international Condor workshop, participating in online forums dedicated to Condor and distributed computing, writing articles and book chapters on using Condor, and delivering invited tutorials at workshops and conferences. "Support" functions serve mainly to deliver new versions of the Condor software. Support also includes maintenance - ongoing bug fixing, support for new operating system releases and new versions of dependent software packages, and updates to documentation. "Enhance" activities will generate basic improvements to the Condor software, such as enhancements in scalability and reliability, new capabilities on Win32, and the incorporation of recent advances in distributed computing technology. "Use" activities leverages the Grid Laboratory of Wisconsin (GLOW) project. GLOW is a multi disciplinary team of researchers across the University of Wisconsin campus that develops, implements, tests and deploys grid-enabled capabilities.Intellectual Merit and Broader Impact - The intellectual merit of the proposed effort lies within the novelty of the distributed mechanisms implemented by Condor and the software engineering challenges the Condor team faces in developing, maintaining and supporting the over 400,000 lines of Condor code in an academic setting. As a leading provider of open-source distributed computing capabilities, Condor plays an important role as a primary "building block" in many research and national and international grid initiatives, including the Grid Physics Network, the International Virtual Data GridLaboratory, the Particle Physics Data Grid, and the NSF Middleware Initiative. Condor serves as the foundation for numerous MS and PhD projects in networking, operating systems, and grid computing. The size, complexity, diversity and quality assurance requirements of the Condor software also lead to challenges in the area of software engineering. The broader impacts of the proposed activity are interdisciplinary and span across both academia and industry. Scientists who use Condor have been able to greatly increase their computing throughput, and therefore increase the size and complexity of the problems they study. Condor is widely used in many disciplines requiring HTC, including Chemistry, Medical, High Energy and Nuclear Physics, Biotechnology, and many others. Condor has been adopted by both universities and government facilities, such as NASA, Sandia National Labs, and NCSA.