A Fair Decentralized Scheduler for Bag-of-Tasks Applications on Desktop Grids

Abstract

Desktop Grids have become very popular nowadays, with projects that include hundred of thousands computers. Desktop grid scheduling faces two challenges. First, the platform is volatile, since users may reclaim their computer at any time, which makes centralized schedulers inappropriate. Second, desktop grids are likely to be shared among several users, thus we must be particularly careful to ensure a fair sharing of the resources.In this paper, we propose a decentralized scheduler for bag-of-tasks applications on desktop grids, which ensures a fair and efficient use of the resources. It aims to provide a similar share of the platform to every application by minimizing their maximum stretch, using completely decentralized algorithms and protocols. After presenting our algorithms, we evaluate them through extensive simulations. We compare our solution to already existing centralized ones under similar conditions, and show that its performance is close to the best centralized algorithms.

Scheduling Tightly-Coupled Applications on Heterogeneous Desktop Grids

“…Most works published in this area are of a pragmatic nature and few theoretical results have been sought or obtained (one exception is the work in [30]). Note that while in this work we focus on scheduling a single application, other authors have studied the scheduling problem for multiple simultaneous applications [31], [32].…”

Decentralized Prioritization-Based Management Systems for Distributed Computing

“…Alternative decentralized approaches to distributed fairshare also exist and include, e.g., solutions for desktop grid systems [6] and bio-inspired algorithms for decentralized load balancing in networks [22]. Decentralization and fair resource scheduling is also an area of interest in related fields, such as wireless systems using non-uniform rate assignment algorithms [23] and studies of information requirements in theoretical economy [5].…”

Bag-of-Tasks applications scheduling on volunteer desktop grids with adaptive information dissemination

“…Celaya et al [12] built the desktop grid as a virtual treebased topology on top of structured P2P overlays. Each tree node is responsible to gather information of its branches and schedule tasks from higher-level nodes, so that both latency and message overhead can be bounded by the network tree height.…”

“…Scheduling concurrent BoT applications is challenging because any delay in completion of any one of the constituent tasks of the bag will affect the overall bag turnaround time. To the best of our knowledge, researchers have only addressed scheduling concurrent bags in centralized desktop grids [6], [7], [8], traditional grid computing [9], [10], [11] or enterprise desktop grids [12], [13]. However the problem of scheduling BoT applications in dynamic and volatile Internet-scale volunteer computing has not been considered.…”