Migrating large output applications to Grid environments: a simple library for threaded transfers with gLite

Abstract

As the EGEE infrastructure and middleware matures, it is becoming increasingly attractive to different kinds of user applications than it was initially planned for. In particular, it starts to attract users of the application paradigm where, from a relatively small input data set, a large set of data is produced, like it is common in several types of simulations. In this paper we present a simple library based on EGEE middleware function calls, in particular lcg utils, that handles the transfer of large data file sets from the local storage of the Worker Node to an EGEE Storage Element in a way that is simple, robust, transparent to the user and minimizes performance impact. The library is very easy to integrate into non-Grid/legacy code and it is flexible enough to allow for overlapping of several data transfers and computation. We also present a carefull analysis of the performance impact on several application scenarios. Finally, we analyse the use of the library in a real case scenario, where it is used in a high-performance plasma simulation code.