Decentralized, Scalable, Grid Fairshare Scheduling (FSGrid)

This work addresses Grid fairshare allocation policy enforcement and presents FSGrid, a decentralized system for Grid-wide fairshare job prioritization. The presented system builds on three contributions; a flexible tree-based policy model that allows delegation of policy definition, a job prioritization algorithm based on local enforcement of distributed fairshare policies, and a decentralized architecture for non-intrusive integration with existing scheduling systems. The system supports organization of users in virtual organizations and divides usage policies into local and global policy components that are defined by resource owners and virtual organizations. The architecture realization is presented in detail along with an evaluation of system behavior in an emulated environment. The system is shown to meet scheduling objectives and convergence noise (mechanisms counteracting policy allocation convergence) are characterized and quantified. System mechanisms are shown to be scalable in tests using realistic policy allocations.