A. A. Bertossi and L. V. Mancini. "Scheduling Algorithms for Fault-Tolerance in Hard-Real-Time Systems". Real-Time Systems, 7(3):229--245, 1994.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....each algorithm s operation relatively to the hardware component on which it is executed. Finally, neither the algorithm to be executed nor the architecture of the system are fixed, but they are inputs of the method. For these reasons, we cannot apply the existing methods, proposed for example in [1, 10, 6, 13, 12], which use preemptive scheduling or approximation methods. Related Work. There exists very little work on this precise topic. Some researchers make hard assumptions on the failure models (e.g. only fail silent, only processor failures) and on the kind of schedule desired (e.g. only static ....

A.A. Bertossi and L.V. Mancini. Scheduling algorithms for fault-tolerance in hard-real-time systems. Real-Time Systems Journal, 7(3):229--245, 1994.

....not satisfy Rtc, then it is the responsibility of the user to add more hardware to increase the redundancy. And second, the obtained schedule must be static to allow optimisations and to minimise the executive overheads. Therefore, we cannot apply the existing methods, proposed for example in [7, 3, 11], which use preemptive scheduling or approximation methods. 2. We want to obtain the schedule automatically, so: The fault tolerance must be obtained without any help from the user. For these two reasons, it will fall into the class of software implemented fault tolerance. 2. Related Work In ....

A. Bertossi and L. Mancini. Scheduling algorithms for fault-tolerance in hard-real-time systems. Real-Time Systems Journal, 7(3):229--245, 1994.

....resulting logic respects unpredictability of faults and can be used to verify fault tolerance. A di erent semantic based approach to verify fault tolerance has been reported in [4] using the Modal Process Logic [16] Finally, fault tolerant scheduling has received quite a lot of attention (e.g. [3], 22] 10] but in a semantic framework which does not give sucient insight into how proofs of other scheduling problems can be carried out, and with synchronisation mostly restricted to precedence between tasks. One direction for our future work is to investigate a formal link between ....

A.A. Bertossi and L.V. Mancini. Scheduling algorithms for fault-tolerance in hard-real-time systems. Real-Time Systems, 7(3):229-245, 1994.

....timing, precedence, and resource constraints of all the tasks can be satisfied, i.e. if the schedulability analysis is successful. Tasks are dispatched according to this feasible schedule. The general problem of optimal fault tolerant scheduling of tasks in a multiprocessor system is NPcomplete [14]. In a real time multiprocessor system, fault tolerance can be provided by scheduling multiple copies of tasks on different processors. Primary backup (PB) and triple modular redundancy (TMR) are two basic approaches that allow multiple copies of a task to be scheduled on different processors. In ....

A.A. Bertossi and L.V. Mancini, "Scheduling algorithms for fault-tolerance in hard-real-time systems, " Real-Time Systems, vol.7, no.3, pp.229-245, 1994.

....instance, i.e. t ij = t ik 1 j k P P i . Furthermore, any scheduling policy which can achieve full utilization on k processors can be used to obtain a feasible schedule with these assignments. An example of such full utilization policies for multiprocessors is provided by Mancini et al. in [23]. We note that the PFair scheduling policy [2] which can also achieve full utilization, assumes that all the periods are multiples of the slot length and hence it can not be used in this context. Corollary 2 Optimal t i values for the Problem MULTI REW can be found by solving the following ....

A. Bertossi and L. V. Mancini. Scheduling Algorithms for Fault-Tolerance in Hard-Real-Time Systems. Real-Time Systems 7:3, p. 229-245, 1994.

....of each algorithm s operation relatively to the hardware component on which it is executed. Finally, neither the algorithm to be executed nor the architecture of the system are xed, but they are inputs of the method. For these reasons, we cannot apply the existing methods, proposed for example in [1, 7, 4, 15, 14], which use preemptive scheduling or approximation methods. There exists a huge amount of work in the domain of fault tolerance of distributed and or real time systems. The following books summarize a part of this work: ffl [28] gives an exhaustive list of the basic concepts and terminology on ....

A.A. Bertossi and L.V. Mancini. Scheduling algorithms for fault-tolerance in hard-real-time systems. Real-Time Systems Journal, 7(3):229245, 1994.

....that is, t ij = t ik 1 j k P P i . Furthermore, any scheduling policy which can achieve full utilization on k processors can be used to obtain a feasible schedule with these assignments. An example of such full utilization policies for multiprocessors is provided by Mancini et al. in [3]. We note that the PFair scheduling policy [1] which can also achieve full utilization, assumes that all the periods are multiples of the slot length and hence it can not be used in this context. Corollary 2 Optimal t i values for the Problem MULTI REW can be found by solving the following ....

A. Bertossi and L. V. Mancini. Scheduling Algorithms for Fault-Tolerance in Hard-Real-Time Systems. Real-Time Systems 7(3): 229-245, 1994.

....to provably tolerate a number of faults, yet be unable to provably tolerate only some of them [12] With unpredictable faults, such fault monotonicity is important but is hard to establish in most branching time theories. Finally, fault tolerant scheduling has recently received some attention [3, 22] but in a semantic framework which does not give sufficient insight into how proofs of feasibility of other scheduling problems can be obtained, and with synchronization restricted to simple precedence between tasks. This paper shows how to realistically analyse the timing properties of ....

A.A. Bertossi and L.V. Mancini. Scheduling algorithms for fault-tolerance in hard-real-time systems. Real-Time Systems, 7(3):229--245, 1994.

....designed either to use periodic checkpoints and reduce the number of processors (see [87] or to use both duplication and checkpoints and maximize the minimum achievable processor utilization (see [82] Both analytical and experimental evaluation of the algorithm performance has been given. In [86] the well known LPT heuristic algorithm for non preemptively scheduling tasks on identical processors has been considered. Backup copies of the tasks have been introduced which have been executed when the computation of the primary copies is lost because of processor failures. Experimental results ....

A. A. Bertossi and L. V. Mancini. Scheduling algorithms for fault-tolerance in hard-real-time systems. Technical Report TR-15/91, Dipartimento di Informatica, Universit`a di Pisa, September 1991. Submitted.

.... short survey, we have not discussed systems that have been built with the aim of providing real time fault tolerance but do not study scheduling specifically [33, 10, 12] For the same reason, we have not discussed scheduling techniques that provide fault tolerance for preemptive real time tasks [26, 2, 22]. 3 The Fault Tolerant Scheduling Problem In this section we describe the system, fault, and task models used in this paper. We also introduce the approach that we have used to schedule dynamic real time tasks with fault tolerance. We consider a system which consists of n identical processors and ....

A A Bertossi and L V Mancini. Scheduling algorithms for fault-tolerance in hard-real-time systems. Technical Report TR-15/91, Univ. of Pisa, Corso Italia 40-56100 Pisa - Italy, 1991. REFERENCES 29

....redundancy is employed to provide a predictable performance in the presence of failures. The response time analysis of [1] is extended to the fault tolerance approach. Fixed point recursive computations are used to determine the response time of each task if faults occur during their execution. In [4], preemptive scheduling algorithms (not using RMS) for fault tolerance in hard real time systems have been developed. This study considers preemptive periodic tasks, and assumes that the system can checkpoint the state of the tasks periodically on a reliable medium in order to handle faults. They ....

A. A. Bertossi and L. V. Mancini, Scheduling Algorithms for Fault-Tolerance in Hard-Real-Time Systems, The Journal of Real-Time Systems, 7(3): 229--245, Nov 1994.

....safety critical systems have active replication to support fault tolerance. However, some benefit could be obtained by allowing a lane to attempt fault recovery without reseting. There are a number of recognised techniques of fault recovery, including recovery blocks [97] and check pointing [98]. Independent of the fault tolerance strategy used, there is still a need to recognise failures using a timing watchdog. The main requirement for the timing watchdog is to identify timing overruns, enabling the functionality for fault tolerance to assess what action is needed. There are two ....

A. Bertossi and L. Mancini, "Scheduling algorithms for fault-tolerance in hard-real-time systems," Real-Time Systems, vol. 7, pp. 229--45, Nov. 1994. Real-Time Syst. (Netherlands).

....two heuristics called Rate Monotonic First Fit (RMFF) and Rate Monotonic Next Fit (RMNF) Recently they were improved upon by Oh and Son [12] The distributed algorithms proposed in [13] address the issue of dynamic scheduling. As for fault tolerant scheduling, some algorithms were presented in [2, 4, 7, 11, 14]. This paper presents the Fault Tolerant DeadlineMonotonic (FTDM) algorithm which extends the RMFF algorithm to a fault tolerant multiprocessor system by using a combined duplication technique where each task scheduled on a processor has either a passive or active copy scheduled on a different ....

A. Bertossi and L. Mancini. Scheduling algorithms for faulttolerance in hard-real-time systems. Real-Time Systems, 7(3):229--245, 1994.

No context found.

A. A. Bertossi and L. V. Mancini. "Scheduling Algorithms for Fault-Tolerance in Hard-Real-Time Systems". Real-Time Systems, 7(3):229--245, 1994.

No context found.

A. Bertossi and L. Mancini. Scheduling algorithms for Fault-Tolerance in Hard Real-Time Systems. Real Time Systems Journal, 7, 1994.

No context found.

A.A. Bertossi and L.V. Mancini, "Scheduling algorithms for fault-tolerance in hard-real-time systems," Real-Time Systems, 7:3, pp.229-245, 1994.

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