D. Haban and K. G. Shin, "Application of Real-Time Monitoring to Scheduling Tasks With Random Execution Times", IEEE Transactions on Software Engineering 16(12) (December 1990).  Home/Search   Document Not in Database   Summary   Related Articles   Check

....since it depends on the actual executions of tasks [8] Spare time may be defined as a situation in which sporadic tasks do not arrive at their maximum rate. Most flexible scheduling algorithms are mainly focused on reclaiming the extra capacity of the system. Only a few research approaches [1, 9, 13] have discussed how to reclaim gain time. Even here, they have tended to focus on procedural programming languages, rather than on object oriented programming languages. This paper introduces a novel gain time reclaiming frame work integrated with WCET analysis to balance the tradeoff among ....

....gives an overview of implementation issues in the real time Java environment and Section 6 evaluates our approach with a practical example. Finally, conclusions and future work are presented in Section 7. 2 Related Work This section gives a brief survey of the related work on gain time analysis [1, 9, 13]. Haban and Shin have proposed an approach [13] placing software triggers at the end of basic blocks in task code to measure actual execution time. By comparing the actual execution time calculated at the software trigger point with pre determined WCET values, the gain time of the specific ....

[Article contains additional citation context not shown here]

D. Haban and K. Shin. Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times. IEEE Transactions on Software Engineering, 16(12), December 1990.

....analysis [10] Many researchers have tackled this problem by focusing on either the implementation of a real time system or the specification of a real time system. Examples of schedulability analysis based on implementations include the works by Stoyenko et al. 10] and Haban and Shin [5]. In [10] a set of language independent schedulability techniques based on the information of program implementation was proposed. In [5] an approach to monitor and verify the task executions was presented. Representatives of schedulability analysis based on specifications include the real time ....

....specification of a real time system. Examples of schedulability analysis based on implementations include the works by Stoyenko et al. 10] and Haban and Shin [5] In [10] a set of language independent schedulability techniques based on the information of program implementation was proposed. In [5], an approach to monitor and verify the task executions was presented. Representatives of schedulability analysis based on specifications include the real time logic technique by Jahanian and Mok [6] and the Petri net based technique by Tsai et al. 11] Our work studies schedulability analysis ....

D. Haban and K.G. Shin, "Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times," IEEE Trans. Software Eng., vol. 16, no. 12, pp. 1374-1389, Dec. 1990.

....since it depends on the actual executions of task [8] The spare time may be defined as a situation in which the sporadic tasks do not arrive at their maximum rate. Most flexible scheduling algorithms are mainly focused on reclaiming the extra capacity of the system. Only a few research approaches [11, 9, 1] have discussed how to reclaim the gain time. However, they have tended to focus on procedural programming languages, rather than on object oriented programming languages. We have proposed a static timing analysis environment for the Java programming language [13] and an approach [12] to address ....

....are given in Section 2 and Section 3 respectively. Section 4 demonstrates how gain time can be reclaimed in objectoriented real time systems. Finally, the conclusion and future work are presented in Section 5. 2 Related Work This section gives a survey of the related work of gain time analysis [11, 9, 1] in the literature. Haban and Shin have proposed an approach [11] placing software triggers at the end of basic blocks in task code to measure actual execution time. In [11] comparing the actual execution time which calculated at the software triggers point with pre determined WCET values, the ....

[Article contains additional citation context not shown here]

D. Haban and K. Shin. Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times. IEEE Transactions on Software Engineering, 16(12), December 1990.

....on the local machine checks for violations as local events happen while sending information about event occurrences needed by remote machines to other remote daemons. Other related work has been reported on the application of real time monitoring for scheduling tasks with random execution times [HaS90]. The main focus of this work is to feed back monitored information to the system for an a posteriori improvement of the behavior. Specifically, the analyzed results about task execution behavior are funneled back to the host s operating system and used for the dynamic scheduling of tasks. The ....

Haban D., Shin K. (1990); "Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times"; IEEE Transactions on Software Engineering; Vol. 16 No. 12 (1990)

....These approaches introduce specialized co processors for the collection and analysis of run time information. The use of special purpose hardware allows for non intrusive monitoring of a system by recording the run time information in a large repository, often for post analysis. In related work [HaS89, TFC90 and HaW90] the use of monitoring information to aid in scheduling tasks has been studied. Here the underutilization of a CPU due to the use of scheduling methods based on the worst case execution times of tasks is approached through a hardware real time monitor that measures the task execution times and ....

Haben D., Shin K.G. (1989); "Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times"; IEEE Real-Time Systems Symposium (1989)

....by ILP, the analysis is likely to become inefficient for larger applications, since solving an ILP takes exponential time. Measurement Based Approaches for Timing Analysis The general idea to use monitoring in real time systems has already been exploited by other researchers. Haban and Shin [Hab90] used it to generate estimates for off line scheduling decisions. This early work has been done on processor architectures that allowed for establishing tight bounds on the WCET by measuring the execution time of the basic blocks of the program. For more advanced architectures, methods known from ....

Haban, D. and K.G. Shin, Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times, IEEE Transactions of Software Engineering, 16(12), pp. 1374-1389, 1990.

....analysis [10] Many researchers have tackled this problem by focusing on either the implementation of a real time system or the specification of a real time system. Examples of schedulability analysis based on implementations include the works by Stoyenko et al. 10] and Haban and Shin [5]. In [10] a set of language independent schedulability techniques based on the information of program implementation was proposed. In [5] an approach to monitor and verify the task executions was presented. Representatives of schedulability analysis based on specifications include the real time ....

....specification of a real time system. Examples of schedulability analysis based on implementations include the works by Stoyenko et al. 10] and Haban and Shin [5] In [10] a set of language independent schedulability techniques based on the information of program implementation was proposed. In [5], an approach to monitor and verify the task executions was presented. Representatives of schedulability analysis based on specifications include the real time logic technique by Jahanian and Mok [6] and the Petri net based technique by Tsai et al. 11] Our work studies schedulability analysis ....

D. Haban and K.G. Shin, "Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times," IEEE Trans. Software Eng., vol. 16, no. 12, pp. 1374-1389, Dec. 1990.

....analysis [10] Many researchers have tackled this problem by focusing on either the implementation of a real time system or the specification of a realtime system. Examples of schedulability analysis based on implementations include the works by Stoyenko et al. [10] and Haban and Shin [5]. In [10] a set of language independent schedulability techniques based on the information of program implementation was proposed. In [5] an approach to monitor and verify the task executions was presented. Representatives of schedulability analysis based on specifications include the real time ....

....the specification of a realtime system. Examples of schedulability analysis based on implementations include the works by Stoyenko et al. [10] and Haban and Shin [5] In [10] a set of language independent schedulability techniques based on the information of program implementation was proposed. In [5], an approach to monitor and verify the task executions was presented. Representatives of schedulability analysis based on specifications include the real time logic technique by Jahanian and Mok [6] and the Petri net based technique by Tsai et al. [11] Our work studies schedulability analysis ....

D. Haban and K. G. Shin, Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times, IEEE Trans. Software Engineering, Vol. 16, No. 12, 1990, pp.1374-1389.

....These experiments are described in the next section. 12.3 Future Work This dissertation is a rst step towards the reusability of software components in real time embedded systems. The usage of a tracker has been explored in synchronous real time systems and real time distributed systems in [25, 26, 27], but no research was found about tracking asynchronous realtime systems. It is clear that the use of an asynchronous model has some interesting bene ts. Nevertheless, a lot of research needs to be done, also experiments with a scheduler placed on top of the tracker (as explained in chapter 9) ....

Haban, D. and K.a.G. Shin, Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times. IEEE Transactions of Software Engineering, 1990. 16(12): p. 1374-1389.

....to derive the mean execution time of a set of processes in the presence of failure [20] there has been little work undertaken to determine the probability distribution of the system s execution time. Much of the research conducted with hard real time systems has focussed on scheduling problems [2, 8, 16, 22, 24], and typically requires knowledge of the execution time bounds of a process to enable efficient schedules to be calculated. With the introduction of fault tolerant procedures, the execution time bounds of the system will change. It is therefore important that a means of deriving an expression for ....

D. Haban and K. G. Shin. Application of real-time monitoring to scheduling tasks with random execution times. IEEE Transactions on software engineering, 16(12), December 1990.

....performance analysis, performance profiling, and job scheduling. The integrated monitor relies on a dynamic combination of probes and sensors, embedded within the VaPoR, Pretsl, and program layers for, respectively, sampling variables and for tracing execution. A job scheduler using milestoning (Haban and Shin, 1989), for example, relies on implanted milestone sensors, which monitor task progress to appropriately reduce the WCET assumed for that task for more efficient online scheduling. All monitored data must be time tagged, collected, filtered, routed, processed, and recorded, while adapting to the dynamic ....

Haban, D. and Shin, K. G. (1989) Application of realtime monitoring to scheduling tasks with random execution times. Proceedings Ninth Real-Time Systems Symposium, December, 1989, 172--181.

.... 1994, Buchmann et al. 1995] which gives a good insight in problems concerning monitoring active (real time) databases [Deutsch, 1994] which has a comparison of different event detection algorithms used in active database; Mellin, 1996] a survey giving an overview event monitoring; [Haban and Shin, 1989, Haban and Wybranietz, 1990, Gorlick, 1991, Plattner, 1984] which concerns event monitoring, in particular hybrid and real time monitoring; Jahanian et al. 1994] which covers hybrid monitoring for scheduling purposes; Snodgrass, 1988] which gives a good overview of high level ....

Haban, D. and Shin, K. G. (1989). Application of realtime monitoring to scheduling tasks with random execution times. In Realtime systems symposium, pages 172--181.

....level is needed, one can decide to turn on additional levels of detail. For this, the source code of the system must be instrumented with sensor code. 5 RELATED WORK The general idea to use monitoring in real time systems has already been exploited by other researchers before. Haban and Shin [HaS 90] used it to generate estimates for scheduling decisions and Raju et al. RRJ 92] describe a system with early timing fault detection. However, our approach to use monitoring information in an on line adaptation loop and for early fault detection is novel. In [MBCW 91, MaW 91] Marzullo and Wood ....

D. Haban, and K.a.G. Shin: "Application of RealTime Monitoring to Scheduling Tasks with Random Execution Times". IEEE Transactions of Software Engineering, 1990. 16(12), pp 1374-1389.

....our on line monitoring system can provide the required generic support for an adaptive scheduler without needing significant additional CPU resources. 4. Related Work The general idea to use monitoring in real time systems has already been exploited by other researchers before. Haban and Shin [5] used it to generate estimates for scheduling decisions and Raju et al. 11] describe a system with early timing fault detection. However, our approach to use monitoring information in an on line adaptation loop and for early fault detection is novel. In [7, 8] Marzullo and Wood propose a ....

Haban, D. and K.a.G. Shin: "Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times". IEEE Transactions of Software Engineering, 1990. 16(12), pp 1374-1389.

....that such margins would be adequate. 1.2 Related Research The research we describe in the remainder of this document has been influenced by two projects. In the first project, Haban and Shin use a real time monitor consisting of dedicated hardware and software to affect the scheduling of tasks [2, 3]. The monitor is installed as a part of the real time system. A task in the system is divided by the programmer into n disjoint partitions based on its structure. These partitions may be classified as Predictable (P) or UnPredictable (UP) depending on how accurately their WETs can be estimated ....

D. Haban and K.G. Shin, "Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times," Proc. the 10th IEEE Real-Time Systems Symposium, pp.172183, 1989.

....[25] These approaches introduce specialized co processors for the collection and analysis of run time information. The use of special purpose hardware allows non intrusive monitoring of a system by recording the run time information in a large repository, often for post analysis. A related work [8] studies the use of monitoring information to aid in scheduling tasks. The underutilization of a CPU due to the use of scheduling methods based on the worst case execution times of tasks is addressed by the use of a hardware real time monitor which measures the task execution times and delays due ....

Haban, D. and Shin, K. G. Dec. 1989. Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times, Proc. IEEE Real-Time Systems Symp., pp. 172-181.

....Source Executable Instrumentation Specification (optional) Instrumentation Library instrumented normal Figure 5: The Building Process using mc4p 5 Related Work The general idea to use monitoring in real time systems has already been exploited by other researchers before. Haban and Shin [6] used it to generate estimates for scheduling decisions and Raju et al. 14] describe a system with early timing fault detection. However, our approach to use monitoring information in an on line adaptation loop and for early fault detection is novel. In [10, 9] Marzullo and Wood propose a ....

Haban, D. and K.a.G. Shin: "Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times". IEEE Transactions of Software Engineering, 1990. 16(12), pp 1374-1389.

.... and Whalley, 1992, Hong and Greber, 1993, Nirkhe and Pugh, 1992) and the run time refinement of WCET estimates based upon a combination of compiletime information (Gupta and Gopinath, 1994, Park and Shaw, 1990) and run time information (Gopinath, Bihari, and Gupta, 1992, Gopinath and Gupta, 1990, Haban and Shin, 1989, Audsley, Davis, and Burns, 1994) Some analysis techniques take into account the architectural characteristics of the system (Lim et al. 1994, Neihaus, 1991) However, we are unaware of any algorithms that automatically generate CTGs or compute busy idle profiles to facilitate interleaved and ....

....Hong s approach requires analyzing the timing behavior of the tasks resulting from the slicing transformation. Our approach for non intrusive monitoring of real time applications can enhance the effectiveness of existing systems (Ogle, Schwan, and Snodgrass, 1993, Tokuda, Kotera, and Mercer, 1988, Haban and Shin, 1989). In contrast to the non intrusive monitoring techniques developed by Tsai, Fang, and Chen (1990) which require specialized hardware support, our approach is based entirely upon software techniques. 6. Concluding Remarks We presented techniques for the timing analysis of a set of communicating ....

Haban, D. and K.G. Shin, "Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times," Proc. 10th IEEE Real-Time Systems Symposium, 1989, pp. 172--183.

....Young Investigator Award CCR 9157371 to the University of Pittsburgh. veniently as the system gets larger[8] It also often results in severe under utilization of the system since tasks typically complete in much less time (sometimes orders of magnitude less) than their WETs would indicate [5]. Another problem with the above approach to scheduling based on worst case execution times arises when a task for some reason, perhaps owing to resource sharing delays, exceeds its WET. This results in deadline failure, but such a failure is noticed very late in the task s lifetime. This reduces ....

....add a safety margin by increasing the WET of a task by some arbitrary percentage. This approach further exacerbates the underutilization problem. In order to address the above problems researchers have proposed run time refinement of execution time estimates based upon monitoring information [4] [5] [7] We had introduced a technique called Compiler Assisted Adaptive Scheduling (CAADS) 4] using which execution times of the various parts of the program are determined at run time. If time savings are observed at run time, they can be used to accommodate newly arriving tasks. On the other ....

D. Haban and K.G. Shin. Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times. Proc. 10th Real-Time Systems Symposium, 1989.

No context found.

D. Haban and K. G. Shin, "Application of Real-Time Monitoring to Scheduling Tasks With Random Execution Times", IEEE Transactions on Software Engineering 16(12) (December 1990).

No context found.

D. Haban and K. Shin. Application of real-time monitoring to scheduling tasks whith random execution times. In IEEE Transaction on software enginneering, December 1990.

No context found.

D. Haban and K. G. Shin. Application of real-time monitoring to scheduling tasks with random execution times. IEEE Transaction on Software Engineering, 16(12), December 1990.

No context found.

Haben D., Shin K., (1990); Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times; IEEE Trans. on Software Engineering; Vol. 16 No. 12.

No context found.

D. Haban and K. G. Shin. Application of real-time monitoring to scheduling tasks with random execution times. IEEE Transaction on Software Engineering, 16(12), December 1990.

No context found.

Haben D., Shin K., (1990). Application of Real-Time Monitoring to Scheduling Tasks with Random Execution Times; IEEE Trans. on Software Engineering; Vol. 6 No. 12.

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