K. Je#ay, D.F. Stanat, and C.U. Martel. On non-preemptive scheduling of periodic and sporadic tasks. In Proc. of the Twelfth IEEE Real-Time Systems Symposium, pages 129--139. IEEE Computer Society Press, 1991.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....system is scheduled using EDF , then the minimum required FLOPs is 112.67. Therefore, we see that for the same multi rate decomposition of the controller, EDF requires 14.54 less computational time than the round robin scheduling algorithm. In general, non preemptive scheduling algorithms [JSM91] insert CPU idle time and hence, require more computational resources to feasibly schedule the same task set than preemptive scheduling algorithms. 109 g g w (rad s) g V w (rad s) g w (rad s) w (rad s) Figure 4.8: Frequency domain tracking response in the ....

K. Je#ay, D. F. Stanat, and C. U. Martel. On Non-Preemptive Scheduling of Periodic and Sporadic Tasks. Proceedings of the 12 IEEE Real-Time Systems Symposium, San Antonio, Texas, IEEE Computer Society Press, pages 129--139, December 1991.

.... control in RT communication in our realtime network RTnet [15] and currently also for Quality of Service control experiments with Bluetooth [16] Another application where we could use the presented technique is Clockwise [17] a mixed media file system, originally based on Je#ay s theory [18], in which the presented scheduling algorithms were used for non preemptable RT disk scheduling. The battle about whether or not to include support for resource sharing in our RT scheduler was won by the resource sharing camp: the schedulability test is straightforward and the run time complexity ....

D. K. Je#ay, Stanat, and C. Martel, "On non-preemptive scheduling of periodic and sporadic tasks," in Proc.ofthe12 th IEEE Real-Time Sys. Symp., 1991, pp. 129--139.

....disk drive. Preemptible disk access is desirable in certain settings. One such domain is that of real time disk scheduling. Real time scheduling theoreticians have developed schedulability tests (the test of whether a task set is schedulable such that all deadlines are met) in various settings [9, 10, 11]. In real time scheduling theory, blocking , or priority inversion, is defined as the time spent when a higher priority task is prevented from running due to the non preemptibility of a low priority task. Blocking degrades schedulability of real time tasks and In this paper, we refer to ....

K. Jeffay, D. F. Stanat, and C. U. Martel. On nonpreemptive scheduling of periodic and sporadic tasks. Proceedings of the Twelfth IEEE Real-Time Systems Symposium, December 1991.

....applicable to general instances and does not make particular assumptions about the parameters. Scan scheduling is also related to other nonpreemptive scheduling problems. In traditional contexts where the objective is to complete all jobs before their deadlines EDF scheduling is optimal [8]. This is no longer true for scan scheduling. For example, selecting bands in decreasing order of the deadlines d i (using the notations of Section 4) does not always work. EDF is not optimal because of the dependency between the start time of one dwell and the deadline for the next dwell. It may ....

K. Jeffay, D. Stanat, and C. Martel. On Non-Preemptive Scheduling of Periodic and Sporadic Tasks. In Proceedings of the 12th IEEE Real-Time Systems Symposium, pages 129-- 139, December 1991.

....code. This makes MTS a hierarchical scheduler. At the top level is ED: if the deadlines of two messages can be distinguished after quantization, then the one with the earlier deadline has higher priority. Non preemptive scheduling under release time constraints is NP hard in the strong sense [13]. However, Zhao and Ramamritham [14] showed that ED performs better than other simple heuristics. 4 deadline DM priority 1 DM priority 0 1 fixed priority 0 0 (a) b) c) Figure 2: Structure of the ID for MTS. Parts (a) through (c) show the IDs for high speed, low speed, and non real time ....

K. Jeffay, D. F. Stanat, and C. U. Martel, "On non-preemptive scheduling of periodic and sporadic tasks," in Proc. Real-Time Systems Symposium, pp. 129--139, 1991.

....Yang Yu # , Shriram B. Gundala # , Sethavidh Gertphol # , Howard Jay Siegel # , Anthony A. Maciejewski # , and Viktor Prasanna # Purdue University # University of Southern California School of Electrical and Computer Engineering Department of Electrical Engineering West Lafayette, IN 47907 1285 USA Los Angeles, CA 90089 2560 USA alis, kim42 ecn.purdue.edu yangyu, gundala, gertphol, prasanna halcyon.usc.edu Colorado State University # Department of Electrical and Computer Engineering Department of Computer Science Fort Collins, CO 80523 1373 USA hj, aam colostate.edu ....

....mapping heuristics designed for the NSWC system. The details of the simulation experiments are given in Section 5. Section 6 concludes the paper. 2. Related Work Many research efforts in the literature concentrate on mapping real time applications on a uniprocessor (e.g. 4, 7, 12, 20, 23, 28, 29, 30, 31, 32, 33, 34, 43, 46] Even though these papers present good schemes to schedule realtime tasks on a uniprocessor, they could not be directly ap sensor actuator non blocking switch Figure 3. The hardware model. All machines have dedicated, full duplex Ethernet connections to a ....

K. Jeffay, D. F. Stanat, and C. U. Martel. On optimal, nonpreemptive scheduling of periodic and sporadic tasks. In 12th Real-Time Systems Symposium, pages 129--139, Dec. 1991.

....preemptively. This means that a schedulability test needs to guarantee that there is sufficient time in the schedule to finish any request that cannot be preempted while also guaranteeing that any other request still meets its deadline. The schedulability test is based on work by Jeffay et al. [53] and is explained later in Section 6.1.2. Table 6.1 lists a number of terms that are used throughout this chapter. The remainder of this chapter is organized as follows. Section 6.1 gives an overview of existing real time scheduling techniques and its applicability to schedule disks. Section 6.2 ....

....a later deadline (i.e. lower priority) It is proven that iff: a task set is feasible. Also, the EDF scheduling algorithm is optimal: if a task set can be scheduled by any dynamic algorithm, it can also be scheduled by an EDF scheduler. 6.1. 2 Nonpreemptive (EDF) scheduling Jeffay et al. [53] have determined the necessary and sufficient set of conditions for a set of sporadic or periodic tasks with arbitrary release times to be schedulable by a nonpreemptive version of Liu and Layland s EDF scheduler. Assume a task set # 1 . # n is sorted in non decreasing order by period. If the ....

[Article contains additional citation context not shown here]

Kevin Jeffay, Donald F. Stanat, and Charles U. Martel. On Non-Preemptive Scheduling on Periodic and Sporadic Tasks. Real-Time Systems Symposium, IEEE TC Real-Time Systems, pages 129--139, 1991.

....applications can execute. described in Section VI, along with the experimental setup and the results for the evaluation of these heuristics. II. Related Work Many research e#orts in the literature concentrate on scheduling real time applications on a processor (e.g. 5] 8] 14] 24] 27] [33], 34] 35] 36] 37] 38] 39] 44] 49] 55] 56] The major ways in which our work di#ers from the works cited above are: a) the system environment of our research is very di#erent from most of the previous research (e.g. multitasking heterogeneous distributed machines capable of ....

K. Je#ay, D. F. Stanat, and C. U. Martel. On optimal, nonpreemptive scheduling of periodic and sporadic tasks. In 12th Real-Time Systems Symposium, pages 129--139, Dec. 1991.

....e ort and with no additional overhead, and it permits to implement the preemption threshold mechanism on top of EDF. The last issue can lead to further optimizations: the EDF scheduling algorithm has been proven optimal both in the preemptive [14, 2, 3] and in the non preemptive 3 version [11]; furthermore, in [13] the authors claim that EDF SRP is an optimal algorithm for scheduling sporadic task sets with shared resources. Since EDF is optimal, it is more likely that a given assignment of preemption thresholds produces a feasible schedule. Therefore, we expect a better chance to ....

K. Jeay, D. F. Stanat, and C. U. Martel. On nonpreemptive scheduling of periodic and sporadic tasks. In Proceedings of the IEEE Real-Time Systems Symposium, pages 129-139, December 1991.

....on the worst case utilization bounds of different deadline postponement strategies. He shows that if one additional period is given to tasks to complete their computation requirement, the worst case schedulable utilization bound increases from 0.690 to 0.811. Jef fay, Stanat, and Martel [7] show a set of scheduling conditions for scheduling periodic tasks and sporadic tasks on a uniprocessor system. Any set of periodic or sporadic tasks that satisfies those conditions can be scheduled with an earliest deadline algorithm. The following sections focus on the scheduling of periodic ....

K. Jeffay, D. F. Stanat, and C. U. Martel, "On non-preemptive scheduling of periodic and sporadic tasks," in Proc. 12th IEEE Real-Time Systems Symp., pp. 129-139, Dec. 1991.

....the case of non preemptive scheduling. Leung [8] considered the non preemptive scheduling prob lem for multiprocessor computing system, where the tasks are of uniform length and where the task execution is restricted according to a given partial order. Jeffay, Stanat and Martel presented in [5] necessary and sufficient conditions for a set of periodic or sporadic tasks to be schedulable by the earliest deadline first algorithm in uniprocessor environment. Khil Maeng and Cho [6] proposed a new algorithm and static scheduling strategy for real time periodic tasks with specified release ....

K. Jeffay, D. F. Stanat and C. U. Martel, "On NonPreemptive scheduling of Periodic and Sporadic Tasks," IEEE Real-Time Systems Symposium, pp. 129-139, 1991.

....efficient feasibility analysis algorithms has been extensively studied, particularly in the context of uniprocessor scheduling. For example, 13, 11, 12, 9, 10, 1, 16] have designed feasibilityanalysis algorithms for uniprocessor systems of periodic tasks in a preemptive environment; in [7, 8, 6], feasibilityanalysis of systems of periodic tasks in nonpreemptive environments is considered. While some of the above research has resulted in the discovery of reasonably efficient feasibility analysis algorithms, many of the results obtained have been negative: some basic and important ....

....for performing feasibility analysis for uniprocessor real time scheduling problems that are provably intractable NP hard in the context of sequential computation. Simulation based feasibility analysis algorithms. Most current feasibility analysis techniques (including the ones described in [13, 11, 10, 2, 12, 1, 8, 16, 15]) are simulation based : given the specification of a task system T and a scheduling algorithm S, we determine whether all hard real time jobs of T will meet al..l deadlines when scheduled by algorithm S by (implicitly or explicitly) simulating the run time behaviour of the system, and observing ....

K. Jeffay, D. Stanat, and C. Martel. On non-preemptive scheduling of periodic and sporadic tasks. In Proceedings of the 12th Real-Time Systems Symposium, pages 129-- 139, San Antonio, Texas, December 1991. IEEE Computer Society Press.

....efficient feasibility analysis algorithms has been extensively studied, particularly in the context of uniprocessor scheduling. For example, 13, 11, 12, 9, 10, 1, 16] have designed feasibilityanalysis algorithms for uniprocessor systems of periodic tasks in a preemptive environment; in [7, 8, 6], feasibilityanalysis of systems of periodic tasks in nonpreemptive environments is considered. While some of the above research has resulted in the discovery of reasonably efficient feasibility analysis algorithms, many of the results obtained have been negative: some basic and important ....

K. Jeffay and R. Anderson. On optimal, non-preemptive scheduling of periodic and sporadic tasks. Technical Report 88-11-06, Department of Computer Science, University of Washington, 1990.

.... [11, 9, 10] have designed scheduling and feasibility analysis algorithms for uniprocessor systems of periodic tasks in a preemptive environment an environment in which a job executing on the processor may be interrupted at any instant and its execution resumed later, with no cost or penalty; in [6], feasibility analysis of systems of periodic tasks in nonpreemptive environments is considered. Scheduling of mixed systems composed of periodic tasks and aperiodic jobs has also been studied; see, e.g. 8, 7] The Chetto Chetto algorithm. Among the more influential bodies of research ....

K. Jeffay and R. Anderson. On optimal, non-preemptive scheduling of periodic and sporadic tasks. Technical report, Department of Computer Science, University of Washington, 1990.

....each task and there are usually multiple tasks in a system, there arises a problem of scheduling all requests of the tasks properly so that their deadlines are met. The development of scheduling algorithms for periodic task systems has been a major focus in the area of real time scheduling theory [1, 16, 19, 20, 24, 28, 40, 42, 43, 46, 47, 60, 64, 68, 70]. Since current technology is incapable of producing hardware components which never fail or software programs which are free of errors, a task might miss its deadline because of processor failures or task errors. To tolerate hardware failures or software errors, hardware and software redundancy ....

....but its performance depends on the pre selection of M and hence S M , where S M is a decreasing function of M, e.g. S M = 2.34 for M = 4, and S M 2.2837 for M . For non preemptive scheduling, the problem of minimizing the number of processors is more difficult. Jeffay, Stanat, and Martel [28] have shown that the problem of determining whether a set of non preemptive, periodic tasks with different release times is schedulable is NP hard in the strong sense. Furthermore, they have shown that a set of periodic tasks may not be schedulable non preemptively on a single processor, even if ....

[Article contains additional citation context not shown here]

K. Jeffay, D.F. Stanat, and C.U. Martel, On non-preemptive scheduling of periodic and sporadic tasks," IEEE Real-Time Systems Symposium, 1991, 129-139.

....periods = 3 8 (sec) D = 100 D = 25 D = 50 Figure 7: Ratio of missed requests versus di erent number of views. After obtaining derived virtual view set V and its trivial candidate query set Q 0 (V ) we apply the traditional EDF (Earliest Deadline First) algorithm [2, 11], which is optimal in most cases, to schedule queries included in Q 0 (V ) and to materialize views in V . Note that if a virtual view v 2 V is integrated from a set of views C v C, then the refresh period of v is determined as the minimum refresh period among all views in C v . 6 Simulation ....

K. Jeffay, D. Stanat, C. Martel. On non-preemptive scheduling of periodic and sporadic tasks. In Proceedings of the 12th IEEE Real-Time Systems Symposium, pages 129-139, 1991.

....is such that disk bandwidth is not a problem. Deadline scheduling [6] is known to be an optimal scheduling strategy when the tasks are preemptable with zero cost and the task completion times are known in advance. Deadline scheduling is shown to be optimal even when the tasks are not preemptable [7]. But, both these studies assume that the task completion times are known in advance. The block transfer time in the network is dependent on whether there is any contention for the switch links and this contention varies based on the network load. Also, a network transfer requires multiple ....

K. Jeffay, D. F. Stanat, and C. U. Martel. On non-preemptive scheduling of periodic and sporadic tasks. Proc. of Real-time Systems Symp., pages 129--139, Dec. 1991.

....of cells in cellbase C to the regular and random cases as we did in Section 3.2 and 3.3. However, unlike the case when tight freshness is required, it is not a simple task to determine the feasibility of scheduling refresh queries in Q(C) when loose freshness is required. It has been shown in [6] that determining the feasibility of a period task set with arbitrary task releasing times and arbitrary task periods is a NP Hard problem. We extend the result here. Theorem 10. Given a cell set C with request pattern set R(C) fhU 1 ; t 1;1 i, hU 2 ; t 2;1 i, hUn ; t n;1 ig, where U i = ....

....for processor time at regular periodic intervals. For non preemptive scheduling, which is our case, work has been done with inserted idle time allowed or not. Without inserted idle time, when a task has been released, it cannot wait before being scheduled if the processor is not busy. As shown in [6], when inserted idle times are not used, feasibility of a periodic task system for all start times can be determined in pseudo polynomial times. They have also shown that determining feasibility for given start times is NP Hard in the strong sense. When inserted idle times are allowed, the problem ....

K. Jeffay, D. Stanat, C. Martel. On non-preemptive scheduling of periodic and sporadic tasks. In Proceedings of the 12th IEEE Real-Time Systems Symposium, pages 129-139, 1991.

....5) However, if Q 0 (C) is scheduled by EDF algorithm, then all cells can be kept loose fresh and no refresh requests are missed. Thus, Q 0 (C) is loose feasible. We say that a scheduling algorithm is global if it can produce feasible schedules for all feasible candidate query sets. According to [11], we know that EDF algorithm is global for candidate query sets. 16 6 r 2;1 6 r 3;1 6 6 r 2;3 6 r1;2 6 r 3;2 6 r2;2 0 2 3 4 5 6 8 1 9 11 10 12 13 14 7 r 1;1 t q 1 q 3 q 2 q 1 Infeasible schedule produced by MaEFT algorithm. r 2;1 has been missed. 6 r 2;1 6 r 3;1 6 6 r ....

K. Jeffay, D. Stanat, C. Martel. On non-preemptive scheduling of periodic and sporadic tasks. In Proceedings of the 12th IEEE Real-Time Systems Symposium, pages 129-139, 1991.

....C with refresh pattern set R(C) fh ; t 1;1 i, h ; t 2;1 i, h ; t n;1 ig, where t 1;1 = 0 and t i;1 , its candidate query set Q(C) fq 1 ; q 2 ; q n g is loose feasible , P n i=1 E q i 6 where E q i is the execution time of query q i . Proof ( The proof follows from [16]. Proof ( To prove the result, we introduce two concepts busy period and idle period . A busy period is a time interval during which the processor is continuously occupied by executions of refresh queries whereas an idle period is a time interval during which processor is idle and no pending ....

....i = k i U 1 (k i 2 Z ) for all 16 i 6n. Although C has a regular view refresh period set, unlike the case when tight freshness is required, it is not a simple task to determine the feasibility of scheduling refresh queries in Q(C) when loose freshness is required. However, it has been shown in [16] that determining the feasibility of a period task set with arbitrary task releasing times and arbitrary task periods is a NP Hard problem. We extend the result here. Theorem 5 Given a view set C with request pattern set R(C) fhU 1 ; t 1;1 i, hU 2 ; t 2;1 i, hU n ; t n;1 ig, where U i = ....

[Article contains additional citation context not shown here]

K. Jeffay, D. Stanat, C. Martel. On non-preemptive scheduling of periodic and sporadic tasks. In Proceedings of the 12th IEEE Real-Time Systems Symposium, pages 129-139, 1991.

.... 10, 11] have designed scheduling and feasibility analysis algorithms for uniprocessor systems of periodic tasks in a preemptive environment an environment in which a job executing on the processor may be interrupted at any instant and its execution resumed later, with no cost or penalty; in [7], feasibility analysis of systems of periodic tasks in nonpreemptive environments is considered. Scheduling of mixed systems composed of periodic tasks and aperiodic jobs has also been studied; see, e.g. 9, 8] The Chetto Chetto algorithm. Among the more influential bodies of research ....

K. Jeffay and R. Anderson. On optimal, non-preemptive scheduling of periodic and sporadic tasks. Technical report, Department of Computer Science, University of Washington, 1990.

....fixed duration tasks or recurring tasks that must be completed within a certain time frame. The problems most studied within the recurring category involve periodically recurring tasks [LL73, LM80, LM81, LW82,Mok83, BHR90] Aperiodically or sporadically recurring tasks have also been studied [Mok83, LSS87, HL88,SLS88, JAM90,SSL89], but currently seem less understood. The main focus of this paper concerns feasibility testing with respect to sporadic task systems (i.e. testing whether a sporadic task system is schedulable) Sporadic tasks were introduced by Mok [Mok83]to model external interrupts# i.e. events external to ....

K. Jeffay, R. Anderson, and C. Martel. On optimal, non-preemptive scheduling of periodic and sporadic tasks. Technical Report TR90 -019, Department of Computer Science, The University of North Carolina at Chapel Hill, April 1990.

....per slot scheduled. The third result is a fast scheduling algorithm based on pinwheel scheduling. 1 Introduction Hard real time systems are systems of tasks where missing a single deadline may have disastrous consequences. Such systems have been the focus of much study over the past twentyyears [13,11,9,12,14,1, 10, 7, 17, 8]. Sporadic tasks in a hard real time environment were introduced by Mok [14] to model a computer responding to external events in a way that is consistent with hard real time systems. They are typically characterized by three positiveintegers e, d, and p, with e d and e p. Two requests for ....

K. Jeffay, R. Anderson, and C. Martel. On optimal, non-preemptivescheduling of periodic and sporadic tasks. Technical Report TR-90-019, Department of Computer Science, The Universityof North Carolina at Chapel Hill, April 1990.

....efficient feasibilityanalysis algorithms has been extensively studied, particularly in the context of uniprocessor scheduling. For example, 13,11, 12, 9, 10,1,16]have designed feasibilityanalysis algorithms for uniprocessor systems of periodic tasks in a preemptive environment# in [7,8, 6], feasibilityanalysis of systems of periodic tasks in nonpreemptive environments is considered. While some of the above research has resulted in the discovery of reasonably efficient feasibility analysis algorithms, many of the results obtained have been negative: some basic and important ....

....for performing feasibility analysis for uniprocessor real time scheduling problems that areprovably intractable NP hard in the context of sequential computation. Simulation based feasibility analysis algorithms. Most current feasibility analysis techniques (including the ones described in [13,11,10,2, 12, 1,8,16, 15]) are simulation based:given the specification of a task system T and a scheduling algorithm S,we determine whether all hard real time jobs of T will meet al..l deadlines when scheduled by algorithm S by (implicitly or explicitly) simulating the run time behaviour of the system, and observing ....

K. Jeffay, D. Stanat, and C. Martel. On non-preemptive scheduling of periodic and sporadic tasks. In Proceedings ofthe12thReal-Time Systems Symposium, pages 129-- 139, San Antonio, Texas, December 1991. IEEE Computer Society Press.

....efficient feasibilityanalysis algorithms has been extensively studied, particularly in the context of uniprocessor scheduling. For example, 13,11, 12, 9, 10,1,16]have designed feasibilityanalysis algorithms for uniprocessor systems of periodic tasks in a preemptive environment# in [7,8, 6], feasibilityanalysis of systems of periodic tasks in nonpreemptive environments is considered. While some of the above research has resulted in the discovery of reasonably efficient feasibility analysis algorithms, many of the results obtained have been negative: some basic and important ....

K. Jeffay and R. Anderson. On optimal, non-preemptive scheduling of periodic and sporadic tasks. Technical Report 88-11-06, Department of Computer Science, UniversityofWashington, 1990.

....In this paper, we present an on line scheduling algorithm for real time systems that attempts to minimize the energy consumed by a periodic task set while also considering the voltage switching times and energies. The algorithm is based on the well known earliest deadline first (EDF) algorithm [9, 7]. We consider a practical scenario where a single CPU executes a set of periodic non preemptable tasks. The voltage, and consequently the clock speed, of the CPU may be switched between two or more values dynamically at run time through OS system calls. This option is available in most modern ....

K. Jeffay, D. F. Stanat and C. U. Martel, "On non-preemptive scheduling of periodic and sporadic tasks with varying execution priority," Proc. IEEE Real-Time Systems Symp., pp. 129--139, December 1991.

....scheduling of tasks with imprecise results. Mok and Dertouzos [27] extended further these results to a multiprocessor environment. Sprunt [32] and Chetto [11] dealt with handling non periodical tasks requests over a set of periodic tasks already scheduled. Non preemptive scheduling is treated by [19]. Resource constraints are handled by [15] 37] Concerning global scheduling, solutions exist both for static tasks placement (some times allowing fault tolerance constraints) 4] 37] and for dynamic task allocation (when task migration is allowed) 9] 34] Moreover, it is known that optimal ....

K.Jeffay, D.Stanat, and C.Martel. On non-preemptive scheduling of periodic and sporadic tasks. In Proceedings of the IEEE Real-Time Systems Symposium, pages 129--129, December 1991.

....that, if any algorithm can schedule a set of tasks, then EDF 6 can scheduled these tasks also. Shared resources were not taken into account yet. Mok [11] has shown that the problem of deciding schedulability of a set of periodic tasks with shared resources (mutex constraints) is NP hard. Jeffay [7] stated that non preemptable EDF can schedule any schedulable task set with shared resources. Audsley et al. 1] presented a necessary and sufficient schedulability test for DM based on interferences, which is related to available computation time. Their technique does not work for EDF. However ....

K. JEFFAY, D.F. STANAT, C.U. MARTEL: "On Non-Preemptive Scheduling of Periodic and Sporadic Tasks," Proc. of the 12 th IEEE Real-Time Sys. Symp., 1991, pp. 129-139.

....not necessarily on meeting specific deadlines, but on providing fair service to all competing streams. It is also easier to think about non preemptive tasks as sporadic tasks, since it has been shown that deciding the schedulability of non preemptive periodic tasks is NP hard in the strong sense [3]. 3. Earliest Deadline First (EDF) scheduling EDF (called deadline driven scheduling in [4] is a popular CPU scheduling algorithm where the priority of a job is based on its deadline. The job with the earliest deadline out of all eligible jobs is assigned the highest priority. In our task model, ....

....execute. Note that in each case, no job misses a deadline. This does not mean that every preemptively schedulable job can be correctly scheduled non preemptively. In fact, the entire problem of deciding whether a periodic task set can be scheduled non preemptively has been shown to be intractable [3]. 4. Fair Queuing and Proportional Share Scheduling Packet scheduling is much like CPU scheduling. Both have elements competing for the use of a shared resource. In the case of CPU scheduling, that resource is the processor, while for packet scheduling, it is the outbound network link. The ....

Jeffay, Kevin, Donald F. Stanat, and Charles U. Martel. On Non-Preemptive Scheduling of Periodic and Sporadic Tasks. Proceedings of the Twelfth IEEE Real-Time Systems Symposium, San Antonio, Texas, December 1991, IEEE Computer Society Press, pp. 129-139.

....[43] Makespan Minimize max. task completion time [45, 47] Lateness Minimize max. task lateness [8, 56] Failure Minimize probability of dynamic failure [21] Schedule Time Time driven dispatching [56] Fixed Fixed priority dispatching [5] Dynamic Dynamic priority dispatching [23] CModel Ideal Zero communication cost Constant Constant communication cost TDC Fixed message priorities [29] KSF Dynamic message priorities [31] Route xy routing Torus chip xy routing [12] Backtrack IncPeriod Increase application period [26] IncReplicate Increase task replication degree ....

K. Jeffay, D. F. Stanat, and C. U. Martel. On non-preemptive scheduling of periodic and sporadic tasks. In Proc. of the IEEE Real-Time Systems Symposium, pp. 129--139, San Antonio, Texas, December 4-- 6, 1991.

.... F taskmodel FFully preemptive We can obtain a more accurate bound (omitted here) for the task model by taking into account the largest blocking time which higher priority tasks will face during execution (which is the size of the largest operation) and modifying the proof found in [7]. We are, however, more concerned with how to improve the feasibility bounds to cope with overloads, through the use of the frequency parameter, f , associated with each operation. Rather than trying to tailor a scheduling algorithm to cater to the frequency parameter (which, if not properly ....

K. Jeffay, D. F. Stanat, and C. U. Mattel. On non-preemptive scheduling of periodic and sporadic tasks. In R. Werner, editor, Proceedings of the Real-Time Systems Symposium - 1991, pages 129--139. IEEE Computer Society Press, December 1991.

....RM method is to allow mutual exclusion. However, the problem of scheduling periodic processes which use semaphores to enforce mutual exclusion is NP hard [Mok83] Another variation is to disallow preemption. Deciding if a set of periodic tasks can be scheduled without preemption is also NP hard [JA88] A schedulability test determines if a set of tasks will all meet their deadlines. The schedulability test may either be static or dynamic. Note that in static schedulers, the generation of a feasible schedule is the schedulability test. As mentioned above, the complexity of deciding ....

K. Jeffay and R. Anderson. On optimal, non-preemptive scheduling of periodic and sporadic tasks. Technical Report 88-11-06, University of Washington, Department of Computer Science, 1988.

....schedulable region have been obtained. However many of these results assume more restrictive arrival patterns than those used in this paper: The optimality of the PEDF for the class of preemptive policies was first shown in [19] for periodic arrivals and in [11] for general arrival patterns; in [16, 17] the delay optimality of NPEDF among the class of non preemptive policies is established for periodic and so called sporadic arrivals; the schedulable regions for NPEDF and PEDF have been derived in [26] for arrival streams characterized by a minimum inter packet arrival time that is independent ....

K. Jeffay, D. F. Stanat, and C. U. Martel. On non-preemptive scheduling of periodic and sporadic tasks. In Proc. Real-Time Systems Symposium, pages 129--139, San Antonio, TX, 1991. IEEE.

....comes closer. 5 IMPLEMENTATION AND TESTS A framework for scheduling experiments with PC and SR EDF has been added to Inferno and is described in (Bos, 97) Test results are not yet available. A considerable amount of work has been done for the determination of the QoS feasibility analysis for PC (Jeffay, 1991) and for SRP EDF (Ripoll, 1996) We could improve the results of the latter by presenting an algorithm with a refined estimation of the blocking component. The details are beyond the scope of this overview, but are ready for publication elsewhere. 6 CONCLUSIONS We have evaluated two dynamic, ....

Jeffay K., Stanat D.F., Martel C.U., (1991) "On Non-Preemptive Scheduling of Periodic and Sporadic Tasks," Proc. of the 12 th IEEE Real-Time Sys. Symp, pp.

....jobs of T i . In the proofs of our scheduling conditions, we also use the notion of task demand , which is related to the notion of unfulfilled demand. The demand placed byataskT i on the processor in an interval [t# t ] is the amount of processing time required byjobsofT i in that interval [45]. In particular, task T i s demand in [t# t ] includes T i s unfulfilled demand at time t, c i time units for each job release of T i in (t# t ] and s time units for eachinterference occurring within (t# t ] in jobs of task T i . A task is said to be inactive at time t if it places 56 ....

K. Jeffay, D. F. Stanat, and C. U. Martel. On non-preemptivescheduling of periodic and sporadic tasks. In Proceedings of the Twelveth IEEE Symposium on Real-Time Systems, pages 129--139. IEEE, December 1991.

....It is common, however, to associate a minimum interarrival time for the instantiations of these unpredictable aperiodic tasks. Much work has been done on scheduling aperiodic tasks with soft deadlines [120] and on aperiodic tasks with hard deadlines, which are known as sporadic tasks [52,111]. Figure 2 4 illustrates an aperiodic task where the arrivals are unpredictable. time rrrr 1234 t h h h time rr 23 h r 1 Current time 12 Figure 2 4: Aperiodic, Unpredictable Task The computation time is another dimension along which tasks may vary. The computation time may be ....

....control algorithm based on a timeline and then used earliest deadline scheduling to actually sequence the tasks. Other systems used analyses from real time scheduling theory to guarantee timing constraints for applications. For example, YARTOS [53] uses algorithms for scheduling sporadic tasks [52] to guarantee timing constraints. In order to guarantee performance, the computational requirements of the applications must be measured and analyzed along with the timing constraints such as delay bounds. Then the application can be run with the expectation that timing constraints will be ....

K. Jeffay, D. F. Stanat, and C. U. Martel. On Non-Preemptive Scheduling of Periodic and Sporadic Tasks. In Proceedings of the IEEE Real-Time Systems Symposium, December 1991.

No context found.

Jeffay, K., Stanat, D. F., and Martel, C. U. 1991. On non-preemptivescheduling of periodic and sporadic tasks. In Proceedings of the 12 th IEEE Symposium on Real-Time Systems #December 1991#, pp. 129#139. IEEE.

No context found.

K. Je#ay, D.F. Stanat, and C.U. Martel. On non-preemptive scheduling of periodic and sporadic tasks. In Proc. of the Twelfth IEEE Real-Time Systems Symposium, pages 129--139. IEEE Computer Society Press, 1991.

No context found.

K. Jeffay, D. F. Stanat, and C. U. Martel, "On NonPreemptive Scheduling of Periodic and Sporadic Tasks", Proc. of the 12th IEEE Real-Time Systems Symposium, San Antonio, TX, December 1991, pp. 129-139

No context found.

K. Jeffay, D. F. Stanat, and C. U. Martel. On non-preemptivescheduling of periodic and sporadic tasks. In Proc. Real-Time Systems Symposium, pages 129--139, San Antonio, TX, 1991. IEEE.

No context found.

K. Jeffay, D. F. Stanat, and C. U. Martel, "On NonPreemptive Scheduling of Periodic and Sporadic Tasks", Proc. of the 12th IEEE Real-Time Systems Symposium, San Antonio, TX, December 1991, pp. 129-139

No context found.

K. Jeffay, D. F. Stanat, and C. U. Martel. On non-preemptive scheduling of periodic and sporadic tasks. In Real-Time Systems Symposium, pages 129--139, 1991.

No context found.

K. Je#ay, D.F. Stanat, and C.U. Martel. On non-preemptive scheduling of periodic and sporadic tasks. In Proc. of the 12th IEEE Real-Time Systems Symposium, pages 129--139. IEEE Computer Society Press, 1991.

No context found.

Jeffay, K., Stanat, D., Martel, C., "On non-preemptive scheduling of periodic and sporadic tasks," Proc. RTSS'91, San Antonio, TX, pp. 129-139, Dec. 1991.

No context found.

K. Je#ay, D. Stanat, and C. Martel. On non-preemptive scheduling of periodic and sporadic tasks. In Proceedings of the 12th Real-Time Systems Symposium, pages 129--139, San Antonio, Texas, December 1991. IEEE Computer Society Press.

No context found.

K. Je#ay, D. F. Stanat, and C. U. Martel. On non-preemptive scheduling of periodic and sporadic tasks. Proceeding of the Twelfth IEEE Real-Time Systems Symposium, pages 129-- 139, 12, 1991.

No context found.

K. Jeffay, D.F. Stanat, and C.U. Martel , On Non-Preemptive Scheduling of Periodic and SporadicTasks , IEEE Real-Time Systems Symposium 1991, pp. 129-139.

No context found.

K. Jeffay, D. F. Stanat and C. U. Martel, "On nonpreemptive scheduling of periodic and sporadic tasks with varying execution priority," Proc. IEEE RealTime Systems Symposium, pp. 129--139, December 1991.

No context found.

K. Jeffay, D. F. Stanat, C. U. Martel, "On Non-Preemptive Scheduling of Periodic and Sporadic Tasks", IEEE Real-Time Systems Symposium, San-Antonio, December 4-6, 1991, pp 129-139.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC