G. Fohler. Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. In Proceedings of the Real-Time Systems Symposium, pages 152--161, 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....hardware overhead costs. On line rescheduling is a hybrid approach where a static schedule is first obtained using an off line scheme. To process aperiodic requests, an on line scheduler shifts the execution of the previously scheduled periodic tasks appropriately without violating their deadlines [10]. This task shifting (b) c) p 2 p b 1 b 2 Figure 3. Example schedules for processor P k : a) Backup schedule S B , b) primary schedule S P , c) schedule S 1 when b 1 is activated, d) schedule S 2 when b 1 and b 2 are activated; e) backup schedule when b 4 and b 5 do not ....

G. Fohler, "Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems," Proc. Real-Time Systems Symp., pp. 152-161, 1995.

....every task to meet its deadline with a specified probability without being interfered with, and greatly simplifies the admission control when combined with the utilization demand analysis. 1. 1 Related Work A number of techniques have been proposed to handle mixes of periodic and aperiodic tasks [13, 16, 6, 17, 7, 8]. The algorithms in [13, 16, 6, 17] assume that aperiodic tasks are soft real time and give preferential treatment to periodic tasks. In these aproaches, aperiodic tasks are handled at a lower priority level in the background, or at a some fixed priority level by a special periodic task which ....

Fohler, G.: Joint Scheduling of Distributed Complex Periodic and Hard Aperiodic Tasks in Statically Scheduled Systems. IEEE Real-Time Systems Symposium, IEEE Computer Society Press (1995), 22--33

....spare capacity or spare laxity to execute aperiodic tasks in any processor on a symmetric multiprocessor system. With this goal in mind, we propose a scheme to allocate aperiodic tasks in multiprocessor hard real time systems. Similar approaches can be found in the context of distributed systems [25,26], but allocating groups of aperiodic tasks and using message passing. These approaches allocate hard aperiodic tasks to nodes where its deadlines will be meet, but they do not specify what to do whenever there are several nodes that can serve the aperiodic tasks. We will focus our study in shared ....

Fohler G., "Joint Scheduling of Distributed Complex Periodic and Hard Aperiodic Tasks in Statically Scheduled Systems", Real-Time Systems Symposium, 152-161, 1995

....straightforward. A common feasible method in multiprocessors systems consist in to partition periodic tasks among processors statically and after to use a well known uniprocessor scheduling algorithm as a local scheduler. Besides, it is also allowed migration of aperiodic tasks to any processor [ 10,11,12]. In this scenario, the periodic task deadlines are guaranteed and aperiodic tasks achieve good response time because the migration algorithm is intended to allocate the task to the most adequate processor. This method is specially interesting for processors with unbalanced loads because migration ....

Fohler G., "Joint Scheduling of Distributed Complex Periodic and Hard Aperiodic Tasks in Statically Scheduled Systems", Real-Time Systems Symposium, 152-161, 1995

....for backup tasks. The technique proposed in [12] requires doubling the number of processors to tolerate a single failure; this level of hardware overhead is unacceptable for many embedded systems. Our approach is somewhat similar to a proposal in [13] which uses flexible scheduling mechanisms from [14] to provide adaptive fault tolerance in statically scheduled systems. The author of [13] uses spare capacities in a static schedule to incorporate various fault tolerant techniques including reexecution. The on line guarantee algorithm presented in [14] does not alter the ordering of tasks imposed ....

....uses flexible scheduling mechanisms from [14] to provide adaptive fault tolerance in statically scheduled systems. The author of [13] uses spare capacities in a static schedule to incorporate various fault tolerant techniques including reexecution. The on line guarantee algorithm presented in [14] does not alter the ordering of tasks imposed by the schedule, whereas the contingency schedules discussed in this paper reorder tasks as long as correctness of the original schedule is preserved. 3 Modeling Assumptions In this section, we describe the system, task, and fault models used in this ....

G. Fohler, "Joint Scheduling of Distributed Complex Periodic and Hard Aperiodic Tasks in Statically Scheduled Systems," Proc. Real-Time Systems Symp., pp. 152-161, 1995.

....data sharing, producer consumer interaction (this is more usual than client server [36] and various forms of precedence orderings. Various studies have show how scheduling parameters can be chosen to produce simple task relationships which nevertheless meet more complex application requirements [29, 30, 26, 10]. Of course a hard activity must never have a synchronous dependency on a non hard job. Similarly a periodic activity cannot rely on the arrival of sporadic or aperiodic jobs. In general, non hard tasks are classified as either firm or soft. A firm task is one that should complete by its deadline ....

....of processor management (scheduling) is to ensure all hard tasks always meet their deadlines and firm tasks are given the maximum resource possible before their deadlines. Table driven cyclic executives support hard tasks but lack the flexibility to support value based scheduling. Although Fohler [26] has shown how some forms of slot movement can enhance the effectiveness of a purely static approach, the approach is not considered further here. Of the two more dynamic schedules, priority based and deadline based, Jorvik relies on preemptive fixed priority (PFP) dispatching. Although in theory ....

G. Fohler. Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. In Proceedings of the 16th Real-Time Systems Symposium, 1995.

....types of constraints can be considered in a preruntime analysis, and an optimal (or sub optimal) solution can be found for the scheduling problem. This method is very powerful, but it is not flexible enough to be applied in dynamic environments, where task arrivals are not known in advance. Fohler [8] proposed a hybrid method for integrating on line service of aperiodic requests with an off line schedule of periodic tasks, but hard and soft tasks cannot share exclusive resources. The same problem has also been considered by a heuristic approach in the Spring kernel [18] but no optimal ....

G. Fohler, Joint Scheduling of Distributed Complex Periodic and Hard Aperiodic Tasks in Statically Scheduled Systems, Proceedings of IEEE Real-Time System Symposium, pp. 152-161, (December 1995).

....among the different tasks. The method in [2] performs resource reclaiming in shared memory realtime multiprocessor systems, where resource reclaiming refers to exploiting a PE at run time when the actual execution time of a task is less than its specified worst case execution time. The work in [23] performs joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. The algorithm in [7] performs concurrent hardware software co synthesis of hard realtime aperiodic and periodic task graphs. This involves allocating the required number of PEs and ....

....have been used in single processor systems, there is only limited work in simultaneously addressing them in distributed systems which execute tasks with precedence constraints. The work presented in [2] only targets resource reclaiming in a multi processor shared memory system. The work in [23] exploits unused resources and leeway in statically scheduled distributed real time systems. As opposed to the work in [23] we perform a global optimization to achieve a better distribution of the flexibility in the static schedule. Moreover, our online scheme is simpler and more powerful in ....

[Article contains additional citation context not shown here]

G. Fohler, "Joint scheduling of distributed complex periodic and hard aperiodic Tasks in statically scheduled systems," in Proc. Realtime Systems Symp., pp. 152-161, Dec. 1995.

No context found.

G. Fohler. Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. In Proc. 16th Real-time Systems Symposium, Pisa, Italy, 1995.

No context found.

G. Fohler (1995). Joint Scheduling of Distributed Complex Periodic and Hard Aperiodic Tasks in Statically Scheduled Systems. In IEEE Real-Time Systems Symposium, December.

No context found.

G. Fohler. Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. In Proceedings of the 16th Real-Time Systems Symposium, Pisa, Italy, Dec. 1995.

No context found.

G. Fohler. Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. In Proceedings of the 16th Real-Time Systems Symposium, Pisa, Italy, Dec. 1995.

No context found.

G. Fohler (1995). Joint Scheduling of Distributed Complex Periodic and Hard Aperiodic Tasks in Statically Scheduled Systems. In IEEE Real-Time Systems Symposium, December.

....computes the overload amount of each aperiodic task. Let 1 : n be a sequence of aperiodic tasks sorted by increasing deadline. Also, assume a sequence of consecutive, non empty, time intervals, each associated to a number of o ine scheduled tasks as de ned by the slotshifting algorithm [Foh95]. The following additional notation is used in the algorithm. dl x the deadline of x c x the remaining execution time of x end x the end time of interval number x sc x the spare capacity of interval number x oa x will be assigned the overload amount of x Algorithm Let ct be the ....

G. Fohler. Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. In Proceedings of the 16th Real-Time Systems Symposium, Pisa, Italy, Dec. 1995. 6

....should be left to the designer, allowing values and importance assigned to tasks, not on their arrival times. In this paper, we present methods for the flexible online handling of firm aperiodic tasks based on offline constructed schedules for time triggered systems. It is based on slot shifting, [4], a method to combine offline and online scheduling by utilizing unused resources. We provide an EDF based acceptance test of O(N) to determine the feasible inclusion of aperiodic tasks into the offline scheduled tasks. The algorithm avoids explicit runtime handling of resource reservations for ....

....systems, e.g. 9] Example algorithms for the selection of tasks to reject in overload situations have been discussed in [3] 7] 2] 1] These algorithms assume control over all tasks in the system and do not take into account the impact of offline scheduled tasks. The algorithm presented in [4] for guarantees of single firm aperiodic tasks on offline schedules does not provide for removal of guaranteed tasks. The rest of the paper is organized as follows: section 2 presents slot shifting, the method we use to combine offline and online scheduling. We introduce the motivation and basic ....

[Article contains additional citation context not shown here]

G. Fohler. Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. In Proc. 16th Real-time Systems Symposium, Pisa, Italy, 1995.

....should be left to the designer, allowing values and importance assigned to tasks, not on their arrival times. In this paper, we present methods for the flexible online handling of firm aperiodic tasks based on offline constructed schedules for time triggered systems. It is based on slot shifting, [4], a method to combine offline and online scheduling by utilizing unused resources. We provide an EDF based acceptance test of O(N) to determine the feasible inclusion of aperiodic tasks into the offline scheduled tasks. The algorithm avoids explicit runtime handling of resource reservations for ....

....systems, e.g. 9] Example algorithms for the selection of tasks to reject in overload situations have been discussed in [3] 7] 2] 1] These algorithms assume control over all tasks in the system and do not take into account the impact of offline scheduled tasks. The algorithm presented in [4] for guarantees of single firm aperiodic tasks on offline schedules does not provide for removal of guaranteed tasks. 2 Offline Complexity Reduction The rigid offline scheduling schemes used for time triggered systems do not provide for flexibility. Including non periodic tasks in the offline ....

[Article contains additional citation context not shown here]

G. Fohler. Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. In Proc. 16th Real-time Systems Symposium, Pisa, Italy, 1995.

....Being part of the SALSART suite the simulator supports distributed cooperative team work, enabling a group of people to work together on a project. 7. 1 Methods Several guarantee algorithms for firm aperiodic tasks based on slot shifting, a method to combine offline and online scheduling [3], are currently supported, including slot shifting for sporadic tasks, 5] overload handling [1] The simulator supports a number of different views; each view showing distinct kinds of information and data to graphically display the inner workings and stages of a simulated algorithm. These ....

G. Fohler. Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. In Proceedings of Real-Time Systems Symposium, Pisa, Italy, Dec. 1995.

....scheduling is very simple as we only compute new deadlines. The use of free resources in offline constructed schedules for aperiodic tasks has been discussed in [11] The resulting flexibility is limited since aperiodic tasks are inserted into the idle times of the schedule only. Slot shifting [7] analysis offline schedules for unused resources and leeway, which is represented as execution intervals and spare capacities. This information is used at runtime to shift task executions, accommodate dynamic tasks, and to perform online guarantee tests. It provides increased flexibility, but ....

....by testing the availability of the specified bandwidth as well. C Transformation technique The feasible schedule with guaranteed bandwidth is transformed into independent tasks with starttimes, deadline pairs. Our method is based on the preparations for online scheduling in slot shifting [7]. The offline scheduler allocates tasks to nodes and resolves the precedence constraints. The scheduling tables list fixed start and end times of task executions, that are less flexible than possible. The only assignments fixed by specification are starts of first and completion of last tasks in ....

G. Fohler. Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. In Proceedings Real-Time Systems Symposium, Dec. 1995.

....describes a runtime scheduling method for a mix of o line scheduled and value based dynamic tasks with a focus on overload situations. We present a formulation of overload handling as a general binary optimisation problem and give an algorithm for solving it. Our method is based on slot shifting [4] to provide for the integration of o ine and online scheduling. Dynamic tasks are scheduled according to basic EDF, extended with an algorithm for overload detection and resolution. We include a penalty value for tasks which have been guaranteed but missed their deadline, e.g. due to rejection ....

....ticks are counted globally, by synchronized clocks with granularity of slot length. Slots have uniform length, and task periods and deadlines must be multiples of the slot length. 2 Task Scheduling and Objectives At run time, scheduling is performed via the slot shifting scheme as detailed in [4] allowing the execution of aperiodic tasks only if this does not cause an o line task to miss its deadline. Upon arrival, aperiodic tasks are inserted into the ready queue of dynamic tasks according to EDF, ensuring optimal performance in non overload situations. There are two basic requirements ....

G. Fohler. Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. In Proc. 16th Real-time Systems Symposium, Pisa, Italy, 1995.

....system lifetime, although overload situations with rejection of less important tasks would be acceptable. 1 This paper describes a runtime scheduling method for a mix of o line scheduled and value based dynamic tasks with a focus on overload situations. Our method is based on slot shifting [Foh95] to provide for the integration of o ine and online scheduling. Dynamic tasks are scheduled according to basic EDF, extended with an algorithm for overload detection and resolution. We include a penalty value for tasks which have been guaranteed but missed their deadline, e.g. due to rejection ....

....that the EDF queue is always free from overload. When a new task arrives, the algorithm tests if it causes overload, and if so, which tasks to reject in order to resolve this eciently. 3.1 Local scheduling algorithm We use the slotshifting method to schedule the tasks on locally on each node. [Foh95] 1 . Due to space limitations, we cannot give a full description here, but con ne to salient features relevant to our new algorithms. More detailed descriptions can be found in [Foh94] IF99] It uses standard o ine schedulers, e.g. Ram90] Foh94] to create schedules which are then ....

[Article contains additional citation context not shown here]

G. Fohler. Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. In Proc. 16th Real-time Systems Symposium, Pisa, Italy, 1995.

No context found.

G. Fohler. Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. In Proceedings of the Real-Time Systems Symposium, pages 152--161, 1995.

No context found.

G. Fohler. Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems. In Proceedings 16th Real-time Systems Symposium, Pisa, Italy, 1995.

No context found.

G. Fohler, "Joint Scheduling of Distributed Complex Periodic and Hard Aperiodic Tasks in Statically Scheduled Systems," Proc. of IEEE Real-Time Systems Symposium, pp. 152-161, Dec. 1995.

No context found.

Gerhard Fohler, "Joint scheduling of distributed complex periodic and hard aperiodic tasks in statically scheduled systems," in Proc. 16-th IEEE Real-Time Systems Symp., Dec. 1995, pp. 152--161.

First 50 documents

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