Introduction to the Theory of Computation
Introduction to the Theory of Computation
Using Run-Time Predictions to Estimate Queue Wait Times and Improve Scheduler Performance
IPPS/SPDP '99/JSSPP '99 Proceedings of the Job Scheduling Strategies for Parallel Processing
SimGrid: A Generic Framework for Large-Scale Distributed Experiments
UKSIM '08 Proceedings of the Tenth International Conference on Computer Modeling and Simulation
Accuracy study and improvement of network simulation in the SimGrid framework
Proceedings of the 2nd International Conference on Simulation Tools and Techniques
Byte-Range Asynchronous Locking in Distributed Settings
PDP '09 Proceedings of the 2009 17th Euromicro International Conference on Parallel, Distributed and Network-based Processing
Iterative computations with ordered read-write locks
Journal of Parallel and Distributed Computing
The parXXL environment: scalable fine grained development for large coarse grained platforms
PARA'06 Proceedings of the 8th international conference on Applied parallel computing: state of the art in scientific computing
Experimenting Iterative Computations with Ordered Read-Write Locks
PDP '10 Proceedings of the 2010 18th Euromicro Conference on Parallel, Distributed and Network-based Processing
| Hi-index | 0.00 |
Data Handover, DHO, is a general purpose API for an efficient management for locking and mapping data. Through objects called lock handles, it enables to control resources in a distributed setting. Such handles ease the access to data for client code, by ensuring data consistency and efficiency at the same time. This paper explores DHO as it was presented in [1]. We model the phases that a lock handle crosses to achieve a DHO locking/mapping life cycle. The Grid Reality And Simulation (GRAS) environment of SimGrid is used as a support of an implementation of DHO and a series of tests and benchmarks of that implementation is presented. GRAS has the advantage of allowing the execution in either the simulator or on a real platform. For that purpose, we exploited a cluster of Grid'5000. The experiments that carried out cover various scenarios of sequences to lock a resources (inclusive or exclusive locking only, or combinations of both) and of combining different architectural factors. The tests demonstrate the ability of DHO to provide a robust and scalable framework. The good evaluation of the present work is consistent with an analysis of the expected behavior done by queuing theory.