Ordonnancement sur plates-formes hétérogènes de tâches partageant des données. (Scheduling Tasks Sharing Data on Heterogeneous Platforms)
نویسنده
چکیده
Résumé Cet article est consacré à l’ordonnancement d’un grand ensemble de tâches indépendantes sur des plates-formes hétérogènes distribuées. Les tâches dépendent de données (en entrée) qui sont initialement réparties sur les différents nœuds de la plate-forme. Une certaine donnée peut être partagée par plusieurs tâches. Pour chaque tâche, notre problème est de décider sur quel nœud de la plate-forme l’exécuter, et de transférer les données nécessaires (celles dont dépend la tâche) vers ce nœud. L’objectif est de trouver une allocation des tâches, et un ordonnancement des communications induites, qui minimisent le temps total d’exécution. Sur le plan théorique, nous exposons différents résultats qui caractérisent la difficulté du problème. Sur le plan pratique, nous proposons plusieurs nouvelles heuristiques, que nous comparons à des heuristiques classiques comme min-min ou sufferage grâce à des simulations.
منابع مشابه
Scheduling Tasks Sharing Files from Distributed Repositories (revised Version) Scheduling Tasks Sharing Files from Distributed Repositories (revised Version)
This paper is devoted to scheduling a large collection of independent tasks onto a large distributed heterogeneous platform, which is composed of a set of servers. Each server is a processor cluster equipped with a file repository. The tasks to be scheduled depend upon (input) files which initially reside on the server repositories. A given file may well be shared by several tasks. For each tas...
متن کاملRestricting EDF migration on uniform heterogeneous multiprocessors
Restricted migration of periodic and sporadic tasks on uniform heterogeneous multiprocessors is considered. Migration between different processors of a multiprocessor causes overhead that may be prohibitively high for real-time systems, where accurate timing is essential. Nonetheless, periodic tasks, which generate jobs at regular intervals, may be able to migrate without causing overhead if th...
متن کاملScheduling Tasks Sharing Files from Distributed Repositories (revised version)
This paper is devoted to scheduling a large collection of independent tasks onto a large distributed heterogeneous platform, which is composed of a set of servers. Each server is a processor cluster equipped with a file repository. The tasks to be scheduled depend upon (input) files which initially reside on the server repositories. A given file may well be shared by several tasks. For each tas...
متن کاملScheduling Trees of Malleable Tasks for Sparse Linear Algebra
Scientific workloads are often described as directed acyclic task graphs. In this paper, we focus on the multifrontal factorization of sparse matrices, whose task graph is structured as a tree of parallel tasks. Among the existing models for parallel tasks, the concept of malleable tasks is especially powerful as it allows each task to be processed on a time-varying number of processors. Follow...
متن کاملEnergy-Aware Scheduling of Flow Applications on Master-Worker Platforms
In this report, we consider the problem of scheduling an application composed of independent tasks on a fully heterogeneous master-worker platform with communication costs. We introduce a bi-criteria approach aiming at maximizing the throughput of the application while minimizing the energy consumed by participating resources. Assuming arbitrary super-linear power consumption laws, we investiga...
متن کامل