Real-time fixed and dynamic priority driven scheduling algorithms: theory and experience
نویسندگان
چکیده
There are two main positions regarding real-time scheduling algorithms. The first is based on fixed priorities and the second makes use of dynamic priorities such as deadlines. These two approaches have never really been compared because the emphasis has always been on the ease of implementation rather than the efficiency of the algorithms and the complexity of the associated feasibility conditions. In addition to traditional real-time applications, we believe that starting to look at these two criteria will be very important from the point of view of providing admission control mechanisms and real-time guarantees on large distributed systems like the Internet network. To that end, our purpose is first to provide a general framework based, on the one hand, a representation of preemptive, real-time scheduling in an algebraic structure that enables us to evaluate the distance of the optimality of any scheduling algorithm ; and on the other hand, a consistent representation of the associated feasibility conditions that enables us to evaluate the number of basic operations. As a second step, considering several kinds of traffics, we initiate the comparison by a straight, but limited, application of our general framework. Our preliminary results will notably highlight, in the cases where deadlines are all greater than periods, that fixed priority schedulers (like deadline monotonic) behave as well as EDF while the worst-case response time analysis is less complex. The same observation is valid when the task sets are almost homogeneous but is in favor of EDF in the general case or when a simple feasibility analysis is needed. Therefore, it might be of interest, given a real-time scheduling context (spanning from small embedded systems to large distributed systems), to take into account these two extra criteria in order to find a right trade-off among several possible solutions. Key-words: busy period, comparison, complexity, efficiency, dynamic priority, fixed priority, preemptive, real-time, scheduling. 1. INRIA, Projet REFLECS, B.P. 105, 78153 LE CHESNAY Cedex (France) Email: {Jean-Francois.Hermant, Nicolas.Rivierre}@inria.fr 2. CNET, France Télécom, PAA/TSA/TLR, 38-40, rue du Général Leclerc, 92794 ISSY-LES-MOULINEAUX Cedex 9 (France) Email: [email protected] Ordonnancement temps réel statique et dynamique: théorie et expérience Résumé : Il existe deux principales familles d’algorithmes d’ordonnancement temps réel, la première s’appuyant sur des priorités fixes et la seconde sur des priorités dynamiques de type échéances. Celles ci n’ont jamais été vraiment comparées l’une à l’autre, si ce n’est en termes de mise en oeuvre. Notre propos est d’initier une telle comparaison en termes d’efficacité des algorithmes ainsi que de complexité des conditions de faisabilité associées. Au delà des applications temps réel traditionnelles, commencer à considérer ces deux critères nous semble devenir critique dans la perspective de la fourniture de mécanismes de contrôles d’admission et de garanties temps réel sur de grands systèmes répartis tels qu’Internet. Dans un premier temps, nous introduisons un cadre général basé, d’une part, sur une représentation de l’ordonnancement temps réel préemptif sous forme de structure algébrique permettant d’évaluer la distance à l’optimalité de tout algorithme et, d’autre part, sur une représentation homogène des conditions de faisabilité associées permettant d’évaluer le nombre d’opérations élémentaires induites. Dans un deuxième temps, considérant différents types de trafics, nous initialisons la comparaison par une application directe, mais limitée, de notre cadre général. Nos résultats préliminaires font notamment apparaître, lorsque les échéances des tâches sont toutes supérieures aux périodes que les algorithmiques d’ordonnancement à priorité fixes (comme Deadline Monotonic) se comportent aussi bien qu’EDF alors que l’analyse des pires temps de réponses des tâches est moins complexe. La même remarque s’applique en présence de tâches homogènes mais reste en faveur de EDF dans le cas le plus général ou lorsqu’une simple analyse de faisabilité est nécessaire. Il semble donc être intéressant, en fonction du contexte d’ordonnancement (allant du petit système embarqué jusqu’au grand système distribué) de faire intervenir les critères d’efficacité et de complexité pour choisir l’algorithme temps réel le plus adapté. Mots-clé : comparaison, complexité, efficacité, période occupée, priorité fixe, priorité dynamique, ordonnancement, préemptif, temps réel. Jean-François Hermant, Laurent Leboucher, Nicolas Rivierre 3
منابع مشابه
Multiprocessor Scheduling For Real Time Systems Embedded Systems
energy-ef?cient tasks scheduling algorithm for real-time utility accrual real-time scheduling for multiprocessor multiprocessor real-time scheduling in industrial embedded a novel approach for off-line multiprocessor scheduling in multiprocessor scheduling for real time systems embedded robust partitioned scheduling for realtime multiprocessor multiprocessor scheduling for real time systems emb...
متن کاملFPCL and FPZL Schedulability Analysis
This paper presents the Fixed Priority until Critical Laxity (FPCL) and Fixed Priority until Zero Laxity (FPZL) and scheduling algorithms for multiprocessor real-time systems. FPZL is similar to global fixed priority pre-emptive scheduling; however, whenever a task reaches a state of zero laxity it is given the highest priority. FPCL is a variant of FPZL that introduces no additional scheduling...
متن کاملAn Annotated Bibliography on Scheduling Algorithms for Fixed and Dynamic Priority Real-Time Systems, with Hard, Firm or Soft Deadline Tasks, and Emphasis on Feasibility Testing [Draft]
In this report we present an annotated bibliography on research work relating to real-time scheduling of tasks. Special emphasis is placed on the different off-line and on-line schedulability testing algorithms for hard, firm and soft real-time tasks. Each proposed algorithm is presented and analyzed in terms of asymptotic time and space complexity. Algorithms are critically examined, when ther...
متن کاملEfficient Dynamic Multilevel Priority Task Scheduling For Wireless Sensor Networks
ABSTRACT: Most of the existing packet-scheduling mechanism of WSN use First Come First Served(FCFS),nonpreemptive priority and preemptive prior it scheduling algorithms. These algorithms incur a high processing overhead and long end-to-end data transmission delay due to improper allocation of data packets to queues in multilevel queue scheduling algorithms. Moreover, these algorithms are not dy...
متن کاملFixed-priority scheduling algorithms with multiple objectives in hard real-time systems
In the context of Fixed-Priority Scheduling in Real-Time Systems, we investigate scheduling mechanisms for supporting systems where, in addition to timing constraints, their performance with respect to additional QoS requirements must be improved. This type of situation may occur when the worst-case resource requirements of all or some running tasks cannot be simultaneously met due to task cont...
متن کاملDynamic Voltage Scaling for Priority-Driven Distributed Real-Time Systems
Energy consumption is increasingly affecting battery life and cooling for computer systems. Dynamic Voltage and frequency Scaling (DVS) has been shown to substantially reduce the amount of power required for uniprocessor and multiprocessor real-time systems that have independent tasks or a statically computed schedule. However, no DVS algorithm has been demonstrated for tasks with precedence co...
متن کامل