Improvements and Study of the Accuracy of the Tasks Duration Predictor, New Heuristics - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Rapport (Rapport De Recherche) Année : 2004

Improvements and Study of the Accuracy of the Tasks Duration Predictor, New Heuristics

Yves Caniou
  • Fonction : Auteur
  • PersonId : 830502
  • IdRef : 084594896
Emmanuel Jeannot

Résumé

The Historical Trace Manager is a task duration predictor module embedded in the agent of a Problem Solving Environment relying on the client-agent-server. The HTM is introduced in and . In this paper, we explain some improvements built into the HTM and NetSolve, the Problem Soving Environment we use for our tests, in order to synchronize the HTM to the reality.We also introduce two new scheduling heuritics relying on the HTM information: Advanced HMCT and Minimum Length.We study the scheduling of several scenarios, including the simultaneous submissions of DAGS and independent tasks, on a real heterogeneous platform.The excellent behavior of the HTM validates its estimations of the duration of each task concurrently running in the system. It can consequently predict the contention tasks may have on each other if scheduled and executed concurrently on the same computing resource.Heuristics performances show the relevancy of the HTM information through the experiments: their ability of behaving with a constant quality between two executions of the same experiment as well as the quality of their respective scheduling choices to optimize several criteria at the same time. We also show that heuristics which rely on minimizing the contention give generally the best results regardless the criterion.We finally compare the behavior of the heuristics previously tested in to the one observed here with more precise information on the global system state due to the synchronization mechanisms. Surprisingly, in the time-shared model, it does not necessarily improve the job repartition among the servers, performances can consequently decrease and the utilization of the fastest servers can become critical.
Fichier principal
Vignette du fichier
RR-5206.pdf (721.75 Ko) Télécharger le fichier
Loading...

Dates et versions

inria-00070786 , version 1 (19-05-2006)

Identifiants

  • HAL Id : inria-00070786 , version 1

Citer

Yves Caniou, Emmanuel Jeannot. Improvements and Study of the Accuracy of the Tasks Duration Predictor, New Heuristics. [Research Report] RR-5206, INRIA. 2004, pp.52. ⟨inria-00070786⟩
95 Consultations
171 Téléchargements

Partager

Gmail Facebook X LinkedIn More