A Grid-enabled Interface to Condor for Interactive Analysis on Handheld and Resource-limited Devices

This paper presents the design and implementation of a Grid-enabled interactive analysis environment for handheld and other resource-limited computing devices. Handheld devices are inherently limited in resources, such as processing power, memory usage, permanent storage and network connectivity. However, handheld devices offer great potential because they provide ubiquitous access to data and round-the-clock connectivity. Our solution aims to provide users of handheld devices the capability to launch heavy computational tasks on computational and data grids, monitor the jobs ’ status during execution, and retrieve results after job completion. A web service-hosting framework named JClarens has been developed to act as the middleware for job scheduling, submission and execution. Users carry their jobs on their handheld devices in the form of executables (and associated libraries). Files to be analyzed remotely are served by JClarens. JClarens then schedules the jobs on the most suitable farm on the Grid, using monitoring information gathered from all the connected farms. It also manages the transfer of executables and selected files from the handheld devices to the selected farm, the submission of the job to the Condor scheduler on that farm and the subsequent monitoring of the jobs ’ status. Thus users can transparently view the status of their jobs and finally get back their outputs without having to know where they are being executed. In this way, our system is able to act as a high-throughput computing environment where devices ranging from powerful desktop machines to small handhelds can employ the power of Grid computing. Keywords: Grid computing, Handheld Computing, Physics Analysis, Context-aware job submission. 1