Throughput Optimization for Pipeline Workflow Scheduling with Setup Times

نویسندگان

  • Anne Benoit
  • Mathias Coqblin
  • Jean-Marc Nicod
  • Laurent Philippe
  • Veronika Rehn-Sonigo
چکیده

In this paper, we tackle pipeline workflow applications that are executed on a distributed platform with setup times. In such applications, several computation stages are interconnected as a linear application graph, and each stage holds a buffer of limited size where intermediate results are stored and a processor setup time occurs when passing from one stage to another. The considered stage/processor mapping strategy is based on interval mappings, where an interval of consecutive stages is performed by the same processor and the objective is the throughput optimization. Typical examples for this kind of applications are streaming applications such as audio and video coding or decoding, image processing using co-processing devices as FPGA. Even when neglecting setup times, the problem is NP-hard on heterogeneous platforms and we therefore restrict to homogeneous resources. We provide an optimal algorithm for constellations with identical buffer capacities. When buffer sizes are not fixed, we deal with the problem of allocating the buffers in shared memory and present a b/(b+ 1)-approximation algorithm. Key-words: setup times; buffer; coarse-grain workflow application; throughput; complexity results. ha l-0 06 74 05 7, v er si on 2 19 J un 2 01 2 Optimisation du débit dans l’ordonnancement des flux en pipeline avec temps de reconfiguration Résumé : Dans ce rapport, nous traitons les applications de type pipeline exécutées sur des platformes distribuées avec des temps de reconfiguration. Dans ce type d’applications, les différentes tâches qui composent un même calcul sont interconnectées selon un graphe linéaire, et à chaque tâche est assigné un buffer de taille limitée, dans lequel les résultats intermédiaires sont stockés. Un temps de reconfiguration est nécessaire à chaque processeur pour passer de l’exécution d’une tâche à une autre. La stratégie d’affectation considérée des tâches sur les processeurs est basée sur le principe d’une allocation par intervalles, dans laquelle un intervalle de tâches consécutives est affecté au même processeur. L’objectif est ici l’optimisation du débit. Un exemple d’applications regroupe celles gérant des flux de données (streaming applications), comme l’encodage/décodage audio et vidéo, ou le traitement d’un flux d’images à l’aide de coprocesseurs tels que le FPGA. De plus, même sans prendre en compte les temps de reconfiguration, le problème est NP-complet dès lors que les processeurs sont hétérogènes, même avec des communications homogènes. Nous proposons un algorithme optimal pour l’ordonnancement des tâches au sein d’un même processeur disposant de buffers de tailles identiques. Lorsque la taille des buffers n’est pas fixée, nous traitons le problème de l’allocation des buffers partageant une même mémoire et proposons une approche heuristique, notamment une b/(b+ 1)-approximation. Mots-clés : temps de reconfiguration; buffer; flux de travaux; débit; résultats de complexité. ha l-0 06 74 05 7, v er si on 2 19 J un 2 01 2 Throughput optimization for pipeline workflow scheduling with setup times 3

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

A New ILP Model for Identical Parallel-Machine Scheduling with Family Setup Times Minimizing the Total Weighted Flow Time by a Genetic Algorithm

This paper presents a novel, integer-linear programming (ILP) model for an identical parallel-machine scheduling problem with family setup times that minimizes the total weighted flow time (TWFT). Some researchers have addressed parallel-machine scheduling problems in the literature over the last three decades. However, the existing studies have been limited to the research of independent jobs,...

متن کامل

Meta heuristic for Minimizing Makespan in a Flow-line Manufacturing Cell with Sequence Dependent Family Setup Times

This paper presents a new mathematical model for the problem of scheduling part families and jobs within each part family in a flow line manufacturing cell where the setup times for each family are sequence dependent and it is desired to minimize the maximum completion time of the last job on the last machine (makespan) while processing parts (jobs) in each family together. Gaining an optimal s...

متن کامل

A comparison of algorithms for minimizing the sum of earliness and tardiness in hybrid flow-shop scheduling problem with unrelated parallel machines and sequence-dependent setup times

In this paper, the flow-shop scheduling problem with unrelated parallel machines at each stage as well as sequence-dependent setup times under minimization of the sum of earliness and tardiness are studied. The processing times, setup times and due-dates are known in advance. To solve the problem, we introduce a hybrid memetic algorithm as well as a particle swarm optimization algorithm combine...

متن کامل

Pareto-based Multi-criteria Evolutionary Algorithm for Parallel Machines Scheduling Problem with Sequence-dependent Setup Times

This paper addresses an unrelated multi-machine scheduling problem with sequence-dependent setup time, release date and processing set restriction to minimize the sum of weighted earliness/tardiness penalties and the sum of completion times, which is known to be NP-hard. A Mixed Integer Programming (MIP) model is proposed to formulate the considered multi-criteria problem. Also, to solve the mo...

متن کامل

Two meta-heuristic algorithms for parallel machines scheduling problem with past-sequence-dependent setup times and effects of deterioration and learning

This paper considers identical parallel machines scheduling problem with past-sequence-dependent setup times, deteriorating jobs and learning effects, in which the actual processing time of a job on each machine is given as a function of the processing times of the jobs already processed and its scheduled position on the corresponding machine. In addition, the setup time of a job on each machin...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2012