A. Burns, S. Punnekkat, L. Strigini, and D. Wright. Probabilistic scheduling guarantees for fault-tolerant real-time systems. Technical Report YCS-311, Department of Computer Science, University of York, 1998.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....aspect of fault tolerance is addressed in fault tolerant scheduling. In this scheduling theory, time for recovery is included in the schedule while still guaranteeing that all tasks meet their deadlines. Fixed priority scheduling is extended to include time for reexecution of failed tasks in [13]. However, it is not necessary to reserve time for re executing all tasks in advance, if one assumes that there is an upper bound on the number of tasks that can be affected by transient faults during a specific time interval. Using several time redundant copies of a task and taking a vote on ....

A. Bums, S. Punnekkat, L. Strigini, and D.R. Wright, "Probabilistic Scheduling Guarantees for Fault-Tolerant Real-Time Systems", in Dependable Computing for Critical Applications 7, Piscataway, NJ, USA, 1999, pp. 361-378.

....the notion of probability used is limited and is based on trusted nodes which are likely to accept a transferred task because of their lighter load or proximity. The paper [9] shares the interest of the given paper in optimal scheduling but is based on Markov decision processes. The paper [4] performs a probabilistic schedulability analysis in conjunction with a fault model for fault tolerant execution of tasks. This paper is structured as follows. Section II introduces the basic elements of the analytical framework. Section III introduces a definition of processor assignment for ....

A. Burns, S. Punnekkat, L. Strigini, and D. R. Wright. Probabilistic scheduling guarantees for fault--tolerant real--time systems. In Seventh IFIP International Conference on Dependable Computing for Critical Applications (DCCA--7), San Jose, California, January 6--8, 1999.

No context found.

A. Burns, S. Punnekkat, L. Strigini, and D. Wright. Probabilistic scheduling guarantees for fault-tolerant real-time systems. Technical Report YCS-311, Department of Computer Science, University of York, 1998.

No context found.

A. Burns, S. Punnekkat, L. Strigini, and D.R. Wright. Probabilistic scheduling guarantees for fault-tolerant real-time systems. In Proceedings of the 7th International Working Conference on Dependable Computing for Critical Applications. San Jose, California, pages 339--356, 1999.

....is not normally considered. However, stochastic execution times in schedulability analysis for rate monotonic scheduling is presented in [16] and extended to both static and dynamic scheduling in [17] Also, probabilistic scheduling guarantees for fault tolerant real time systems is considered in [3]. In contrast with schedulability analysis, reliability modelling involves study of fault models, characterisation of distribution functions of faults and development of methods and tools for composing these models and distributions in estimating an overall reliability figure for the syster We ....

A. Bums, S. Punnekkat, L. Strigini, and D.R. Wright. Probabilistic scheduling guarantees for fault-tolerant realtime systems. Proceedings of DCCS-7,IFIP International Working Conference on Dependable Computing for Critical Applications, Califomia, January 1999.

.... because in the worst case r fails just before its completion) The longest interference over ri happens when C is maximum among all tasks in hpe(i) Therefore, in the worst case, the response time can be given by t i C i q y Cj q F rChpe(i) rj Chp(i) 2) The above equation, derived in [3, 4], is valid only when primary tasks and their respective alternative tasks run with the same priority. However, it would be advantageous in certain situations to give higher priorities to alternative tasks since a faulty task certainly has less time to meet its deadline. As an illustration, ....

A. Burns, S. Punnekkat, L. Stringini, and D. Wright. Probabilistic scheduling guarantees for fault-tolerant real-time systems. In Proc. of the 7th International Working Conference on Dependable Computing for Critical Application, pages 339-356, California, USA, Jan 1999.

No context found.

A. Bums, S. Punnekkat, L. Strigini, and D. Wright. Probabilistic scheduling guarantees for fault-tolerant realtime systems. Proceedings of DCCS-7, IFIP International Working Conference on Dependable Computing for Critical Applications, Califomia, January 1999.

....stochastic models, whereas our interest in this paper is primarily on response time analysis using deterministic models, i.e. models in which a bounded worst case scenario is characterized. Studies based on statistical modelling of fault occurrences are useful in providing probabilistic guarantees[3] for timely transfer of message sets. We believe that such an analysis will be a step towards future design of adaptive scheduling strategies which takes in to account the error occurrences and decides online issues such as graceful degradation and choosing di erent policies for di erent classes ....

A. Burns, S. Punnekkat, L. Strigini, and D.R. Wright. Probabilistic scheduling guarantees for fault-tolerant real-time systems. Proceedings of DCCS-7,IFIP International Working Conference on Dependable Computing for Critical Applications, California, January 1999.

....system will meet their deadlines or not 2 [2] 1] 6] 10] The essence of the analysis is to investigate if the deadlines are met in a worst case scenario. Whether this worst case actually will occur during execution, or if it is likely to occur, is not normally considered (an exception being [3]) Reliability modeling, on the other hand involves study of fault models, characterization of distribution functions of faults and development of methods and tools for composing these distributions and models in estimating an overall reliability gure for the system. We have recently [5] ....

A. Burns, S. Punnekkat, L. Strigini, and D.R. Wright. Probabilistic scheduling guarantees for fault-tolerant real-time systems. Proceedings of DCCS-7,IFIP International Working Conference on Dependable Computing for Critical Applications, California, January 1999.

....research community is on hard real time systems, the essence of analysing such systems is to investigate if deadlines are met in a worst case scenario. Whether this worst case actually will occur during execution, or if it is likely to occur, is not normally considered (an exception being [3]) Reliability modeling, on the other hand, involves study of fault models, characterization of distribution functions of faults and development of methods and tools for composing these distributions and models in estimating an overall reliability figure for the system. We have recently [5] ....

A. Burns, S. Punnekkat, L. Strigini, and D. Wright. Probabilistic scheduling guarantees for fault-tolerant real-time systems. Proceedings of DCCS-7,IFIP International Working Conference on Dependable Computing for Critical Applications, California, January 1999.

....system will meet their deadlines or not [4] 5] 8] 11] The essence of the analysis is to investigate if the deadlines are met in a worst case scenario. Whether this worst case actually will occur during execution, or if it is likely to occur, is not normally considered (an exception being [6]) 2 We have recently [7] developed a model for calculating worst case latencies of messages under error assumptions, on the Controller Area Network (CAN) which is a popular and predictable communication network extensively used in the automotive industry and elsewhere. This analysis might ....

A. Burns, S. Punnekkat, L. Strigini, and D.R. Wright. Probabilistic scheduling guarantees for fault-tolerant real-time systems. Proceedings of DCCS-7,IFIP International Working Conference onDependable Computing for Critical Applications, California, January 1999.

....we now first fix u with 0u1 and ask the question What is the probability, P say, that no two of these points are closer than u (conditionally given n) We can obtain the answer by n dimensional integration. This is reported in an extended version of this paper available as a technical report [3], which says essentially that P is just the n th power of the total amount of slack remaining within the unit interval after our u separation constraint is imposed. This solution conditional given n enables us to complete the exact derivation of the final, unconditional P r(W T F ) relatively ....

....as the upper limit of a sum denotes a sum to infinity in the usual sense of a mathematical limit. The purpose of stating this last point about the argument domain now to be assumed for this function is related to the practical computation problem associated with (14) which we address briefly in [3]. Note that, apart from this T F =0 case, the expression (14) represents a finite sum throughout the domain identified, although, for certain argument values, the number of terms summed can be astronomically large, which can make a simple minded numerical computation rather slow. Moreover, some of ....

[Article contains additional citation context not shown here]

A. Burns, S. Punnekkat, L. Strigini, and D.R. Wright. Probabilistic scheduling guarantees for fault-tolerant real-time systems. Technical Report YCS.311, Department of Computer Science, University of York, 1998.

No context found.

A. Burns, S. Punnekkat, L. Stringini, and D. Wright. "Probabilistic Scheduling Guarantees for Fault-Tolerant Real-Time Systems". In Proc. of the 7th Int'l Working Conference on Dependable Computing for Critical Application, pages 339--356, 1999.

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