Real Time Performance Evaluation for Dynamic Scheduling Algorithm

Lei Huang

1998-01-01

Journal of Applied Sciences

Abstract:On the basis of the dynamic scheduling algorithm of real time multitasking and the factors that be closely concerned with real time performance, a evaluation model is proposed. The model can point out that which tasks will produce time out fault and estimate the probability.

What problem does this paper attempt to address?

An Efficient Fault-Tolerant Scheduling Algorithm for Periodic Real-Time Tasks in Heterogeneous Platforms

Weiwei Qiu,Zibin Zheng,Xinyu Wang,Xiaohu Yang

DOI: https://doi.org/10.1109/isorc.2013.6913213

2013-01-01

Abstract:Fault-tolerant real-time scheduling algorithm is one of the most important means to ensure the timeliness and high availability characteristics of fault-tolerant real-time systems. Existing scheduling models for periodic real-time task in heterogeneous platforms typically require the number of processors in the systems to be determined in advance; hence prohibit the scalability and the performance of distributed systems. The algorithms based on these models also require a large number of schedubility tests which lead to long execution time. To address these problems, we propose a primary and backup replica partition based fault-tolerant scheduling algorithm (PBPFT) based on a scalable scheduling model using heterogeneity that does not have to determine the scale of the distributed system in advance. The PBPFT approach also takes advantage of backup copy overlapping and phasing delay techniques to minimize system redundancy, and adopts the processor grouping technique to simplify algorithm complexity. Comprehensive experiments are conducted, and the results validate high resource utilization and commendable performance of our proposed approach.
A Reliability Model for RealTime Multitasking Software

Lei Huang

1998-01-01

Journal of Applied Sciences

Abstract:By taking the task modulars as a basic testing unit, an evaluation model of the realtime multitasking software is presented in this paper. Since the parameters in the model are easy to be analysed and tested, this model is useful in evaluating the reliability of the realtime multitasking software.
Research on Real-Time Scheduling Strategy for Transient Fault Tolerance in Nc System

Xuezeng Pan,Xinhua Yao,Lingdi Ping

DOI: https://doi.org/10.1109/isda.2006.253919

2006-01-01

Abstract:Numerical control (NC) system is a hybrid-tasks system (HTS), which is made up of periodic real-time tasks, aperiodic real-time tasks and non-real-time tasks. To guarantee the time determin-ability and reliability of NC system, a HTS model describing all three kinks of task with temporal parameters tuple was established and the heuristic best-first search (BFS) algorithm was introduced to schedule tasks of this model. Then a fault-tolerance scheduling strategy based on rollback recovery and BFS algorithm was proposed to remove transient faults of system. Comparing to traditional fault-tolerance scheduling algorithm, the strategy presented in this paper is applied to schedule mixed and dependent tasks of NC system, and raises the level of resources utilization by reducing the cost of fault disposal. The validity of this proposed approach applied to NC system is verified in case study and algorithm simulation.
Control Oriented Direct Feedback Scheduling

Feng Xia,Xiaohua Dai,Youxian Sun,Jianxia Shou

2008-01-01

Abstract:The performance of multitasking control systems with resource constraints depends not only on controller design, but also on efficient scheduling of the shared computing resources. To sup- port codesign of feedback control and real-time scheduling under overload conditions, a direct feedback scheduling approach is proposed. The goal is to optimize the overall system perform- ance through dynamically allocating limited CPU time based on control requirements. A con- trol application oriented task model is suggested, providing an interface between control per- formance and task scheduling decision. Based on this model, a novel scheduling algorithm intended for control tasks is presented. The overload condition is effectively handled using a job skipping scheme. Preliminary simulations exhibit the benefits of the proposed approach,
Dynamic Task Scheduler for Real Time Requirement in Cloud Computing System.

Yujie Huang,Quan Zhang,Yujie Cai,Ming-e Jing,Yibo Fan,Xiaoyang Zeng

DOI: https://doi.org/10.1007/978-3-030-05063-4_1

2018-01-01

Abstract:In such an era of big data, the number of tasks submitted to cloud computing system becomes huge and users’ demand for real time has increased. But the existing algorithms rarely take real time into consideration and most of them are static scheduling algorithms. As a result, we ensure real time of cloud computing system under the premise of not influencing the performance on makespan and load balance by proposing a dynamic scheduler called Real Time Dynamic Max-min-min (RTDM) which takes real time, makespan, and load balance into consideration. RTDM is made up of dynamic sequencer and static scheduler. In dynamic sequencer, the tasks are sorted dynamically based on their waiting and execution times to decrease makespan and improve real time. The tasks fetched from the dynamic sequencer to the static scheduler can be seen as static tasks, so we propose an algorithm named Max-min-min in static scheduler which achieves good performance on waiting time, makespan and load balance simultaneously. Experiment results demonstrate that the proposed scheduler greatly improves the performance on real time and makespan compared with the static scheduling algorithms like Max-min, Min-min and PSO, and improves performance on makespan and real time by 1.66% and 17.19% respectively compared to First Come First Serve (FCFS).
Real-time scheduling for tasks with uncertain execution times

Huayong WANG,Yiqi Dai

DOI: https://doi.org/10.3321/j.issn:1000-0054.2005.07.024

2005-01-01

Abstract:A time prediction method was developed to predict the execution time of real-time tasks during execution as the basis for a real-time scheduling system for tasks with uncertain execution times. Comparison of three time prediction models, a VSM-based model, a Markov-based model, and an MVSM-based model, showed that the MVSM-based model most effectively guarantees CPU utilization and the stability of the real-time system. Even with overloads, the scheduling system automatically adjusts, with more than 99% of the tasks completed before their deadlines. The time prediction method resolves the problem of uncertain execution times and provides a reference solution for real-time scheduling systems in uncertain environments.
Research on Dynamic Time Evaluation of Service Procedure and Task Scheduling

Bing Li,Tao Zeng,Xiying Chen,Xiu Li

DOI: https://doi.org/10.1109/CSSE.2008.81

2008-01-01

Abstract:The management of service procedure is the weak point in the service management. This paper proposed a service execution time evaluation method based on the historical information, and the service task dynamic distribution method, to improve the management quality in service procedure. The application condition of these methods, the design of the algorithms and the executive steps are fully investigated.
Survey of Weakly-Hard Real Time Schedule Theory and Its Application

Z Wang,YQ Song,EM Poggi,YX Sun

2002-01-01

Abstract:Normally, tasks are classified into real time and non real time according to temporal constraints for the processing and transmitting of these tasks, consequently the worst-case response time and average performance should be focused on them. However, in practical engineering context, partly violated temporal constraints can be tolerated if the violation meets certain distribution: Nevertheless, the loss-rate (within real time region, an instance of a task is regarded as loss if it violates its temporal constraint) under stable state or statistical real time can solve the problem in some extent, it can't include the permitted distribution of violation. For completely solving the problem, weakly-hard real time schedule theory or window-constraint real time schedule theory, which is used to investigate the problem related to allowing violation of instances over a finite range, consecutive instances or a time window, is proposed. In order to effectively utilize the fact that a practical application can tolerate some violations of temporal constraint under certain distribution, the fundamental research must be done from the aspects of specification of temporal constraint, schedule and schedulibility, and implementation, which are explained in detail in this paper.
A Decomposition-Based Approach for Multitask Scheduling with Execution Uncertainty in Industrial Internet of Things

Yong Wang,Bingtao Hu,Yixiong Feng,Zhiwu Li,Yiping Feng,Jianrong Tan

DOI: https://doi.org/10.1109/jiot.2023.3237727

IF: 10.6

2023-01-01

IEEE Internet of Things Journal

Abstract:Industrial Internet of Things (IIoT) is changing the way in which factories operate with the help of various industrial applications. However, the execution uncertainty of computing tasks has always been ignored in IIoT applications. In this article, we define a novel conditional task graph to describe the execution uncertainty and present a generation algorithm to obtain all task scenario graphs and corresponding occurrence probabilities. Then, a new IIoT-oriented multitask scheduling model under execution uncertainty is built. This model is simplified by reformulating the nonlinear constraints and subsequently decomposed into several small-scale models using the Lagrange multipliers, from which a decomposition-based algorithm is derived to solve the decomposed small-scale models and progressively acquire a well-optimized solution of the initial model. Furthermore, a patching algorithm is constructed to improve the obtained solution. Finally, many test cases are generated, and four selected algorithms are taken for comparison to evaluate the performance of our algorithms. The results demonstrate that our algorithms remarkably outperform the others. Besides, the solutions of the proposed algorithms can completely satisfy the execution deadline constraints of different task scenarios.
Efficient Scheduling Algorithm for Hard Real-Time Tasks in Primary-Backup Based Multiprocessor Systems

王健,孙建伶,王新宇,杨小虎,王申康,陈俊波

DOI: https://doi.org/10.3724/sp.j.1001.2009.00577

2009-01-01

Journal of Software

Abstract:This paper has considered the problem of preemptively scheduling a set of independent periodic hard real-time tasks in primary-backup based multiprocessor systems. An efficient scheduling algorithm—Task Partition based Fault Tolerant Rate-Monotonic (TPFTRM) is proposed which extends RM algorithm to primary-backup based multiprocessor to provide fault tolerance. Compared with previous scheduling algorithms in this area, TPFTRM abandons active backup copies and only uses passive and overlapping backup copies to maximize the backup over-booking and deallocation, thus improves the scheduling performance. Moreover, TPFTRM proposes the task partitioning and processors grouping technique, which reduce the scheduling computation time and also make an easy way to understand and implement it. Extensive simulations experiments are also carried out based on task sets with different parameters. And the simulation result shows a remarkable saving of processors as well as scheduling computation time compared with previous algorithms, which proves the feasibility and effectiveness of the proposed
Time Analysis for Fault-Tolerant Real-Time Group-Based Scheduling

W Lei,ZH Wu,WX Ming

DOI: https://doi.org/10.1109/rams.2005.1408341

2005-01-01

Abstract:The group-based preemptive scheduling provides a flexible mechanism to define the preemptive relations between tasks. However, this scheduling scheme together with a resource access synchronization protocol and requirements of fault tolerance makes the predication of a real-time system's behaviors more difficult than traditional scheduling scheme. The major contribution of our work is an algorithm to calculate the worst-case response time for tasks under the fault-tolerant group-based preemptive scheduling. We first present the algorithm assuming that no task suffers any error. Then the algorithm is extended to consider a fault model and a primary/alternative fault-tolerant mechanism. Based on the tasks' worst-case response time, the schedulability of a given task set can be determined. Moreover, we also introduce a method to calculate the minimum allowed time between two consecutive faults.
A Real-Time Scheduling Algorithm for Embedded Systems with Various Resource Requirements

Chen Tianzhou,Hu Wei,Xie Bin,Yan Like

DOI: https://doi.org/10.1109/iwnas.2006.12

2006-01-01

Abstract:The Scheduling algorithm is one of the most important portions of embedded operating systems especially for real-time embedded operating systems. The performance of scheduling algorithm will influence the performance of the whole system. Real-time embedded operating system needs better response time for real-time process; it is more rigid on response time for hard real-time embedded operating system. A high-performance scheduling algorithm is required in order to make embedded operating system more efficient. Most of the mainstream scheduling algorithms are able to ensure response time for most processes, but for specific real-time applications for embedded system, a specialized real-time scheduling algorithm is required to ensure real-time response. In this paper, we propose a novel real-time scheduling algorithm for real-time embedded system. This algorithm divides all tasks into different Sets and schedules the tasks according to the features of the Set.
Partial preemptive real-time scheduling algorithm in software fault-tolerant model

王健,孙建伶,王新宇,杨小虎,王申康

DOI: https://doi.org/10.3785/j.issn.1008-973X.2009.06.014

2009-01-01

Abstract:A scheduling algorithm called partial-preemptive prediction algorithm (PPA) was proposed in the software fault-tolerant model to reduce the preemption. PPA considers how to reduce the preemption during scheduling as well as how to execute the primaries as many as possible. PPA uses a method similar to non-preemptive earliest deadline first-algorithm(EDFA) to schedule the primaries. Simulations on task sets with different CPU utilization and fault possibility show that PPA can obtain the similar scheduling performance as well as the well-known algorithms so far. Moreover, PPA can reduce the preemption dramatically than previous algorithms under certain conditions, thus reduces the negative impact introduced by the preemption such as overhead runtime computation time.
Study and Implementation of Real-Time Scheduling Mechanism Based on Linux Kernel

Shenglin Ma

2008-01-01

Computer Engineering and Applications Journal

Abstract:In the real time system,the task scheduling strategy is the important part of the kernel design.How to process the task scheduling and make sure the task can complete is the important domain in the real-time operating system research.According to the shortcoming in view of the RT-LINUX scheduler when the system near the load or the overload which the performance suddenly drops,an adaptive value-based scheduler is proposed to achieve a high system value by the less deadlines miss.It can be evaluated the performance of the proposed scheduler in terms of two performance metrics,namely,value ratio and success ratio through both simulation and implementation.
Fault-tolerant real-time tasks scheduling with dynamic fault handling

Gang Chen,Nan Guan,Kai Huang,Wang Yi

DOI: https://doi.org/10.1016/j.sysarc.2019.101688

IF: 5.836

2020-01-01

Journal of Systems Architecture

Abstract:Predictable performance when coping with transient failures is of paramount importance in safety-critical real-time systems. Various software fault-tolerant techniques are employed towards this goal among which check-pointing is a relatively cost-effective scheme. In this paper, we propose an efficient fault-tolerant scheduling framework with run-time fault handling protocol, where criticality levels can be adaptively inserted for fault handling according to run-time fault workload. In contrast to prior works which apply with task re-execution strategy, the proposed framework adaptively determines on-demand re-executions only on the faulty checkpoint segments, rather than on the whole job. Towards this, a unified overrun handling protocol is developed to handle fault recovery adaptively to avoid over-provisioning of resources. In addition, we develop an off-line schedulability analysis technique for the proposed scheduling algorithm. The simulation results show that our fault-tolerant scheduling framework can bring up to 81% improvement in supporting low-criticality service without sacrifice in the MC-schedulability compared with the existing techniques.

computer science, software engineering, hardware & architecture
Reliable Estimation of Execution Time of MapReduce Program

Yang Xiao,Sun Jianling

2011-01-01

China Communications

Abstract:As data volume grows, many enterprises are considering using MapReduce for its simplicity. However, how to evaluate the performance improvement before deployment is still an issue. Current researches of MapReduce performance are mainly based on monitoring and simulation, and lack mathematical models. In this paper, we present a simple but powerful performance model for the prediction of the execution time of a MapReduce program with limited resources. We study each component of MapReduce framework, and analyze the relation between the overall performance and the number of mappers and reducers based on our model. Two typical MapReduce programs are evaluated in a small cluster with 13 nodes. Experimental results show that the mathematical performance model can estimate the execution time of MapReduce programs reliably. According to our model, number of mappers and reducers can be tuned to form a better execution pipeline and lead to better performance. The model also points out potential bottlenecks of the framework and future improvement.
Research on Dynamic Feedback and Elastic Scheduling Model and Algorithm for Flexible Workload

Y Chen,QH Dai

DOI: https://doi.org/10.1109/iccnmc.2003.1243057

2003-01-01

Abstract:Soft real-time applications, such as multimedia systems, are becoming increasingly unpredictable due to the variation of the tasks' attributes. Under these circumstances, the scheduling algorithms depending on the tasks' static attributes can't provide useful and efficient scheduling support to those soft real-time systems. In this paper, we present a dynamic feedback and elastic scheduling model and period adjustment algorithm for soft real-time application with flexible workload. Based on logging the number of task instance that miss deadline periodically, this model adjusts the tasks' executing period to change the resource requirement in the next sampling period. This model makes the system state converge to the steady state when the system is overloaded, and makes tasks to use system resource sufficiently when system resource isn't fully utilized. We analyse the model and evaluate its performance.
Performance evaluation of a dynamically adaptive co-scheduling algorithm

XiaoBo Zhou,Shanglu Lu,Li Xie

1998-01-01

Ruan Jian Xue Bao/Journal of Software

Abstract:The performance evaluation of a dynamically adaptive co-scheduling algorithm - Dasic based on network of workstations is presented. In terms of the aspects of mean response time and system flow, the Dasic algorithm is compared with traditional co-scheduling algorithm MAX and Grab algorithm by simulation. The conclusion is that dynamic adaptation can strikingly improve co-scheduling performance in a massive system of network of workstations.
A New Energy-Efficient Online Scheduling Algorithm for Static Priority Real-Time Systems

LEI Ting,HU Xiao,ZHOU Xue-hai

DOI: https://doi.org/10.3969/j.issn.0253-2778.2006.02.018

2006-01-01

Abstract:Power is a valuable resource in embedded real-time systems as the lifetime of many such systems is constrained by their battery capacity.Recent advances in processor design have added support for dynamic frequency/voltage scaling for saving power and energy.Static priority scheduling algorithms is widely used in real-time systems and energy-efficient scheduling algorithms for static priority real-time systems are urgently needed to be designed. The limitations of energy-efficient scheduling were discussed and a new energy-efficient voltage scaling algorithm was proposed based on the rate monotonic algorithms.The algorithm can analyze slack time more effectively and try to balance the distribution of slack time among tasks of different priorities.A two-phase frequency scaling strategy was designed in order to change the execution time of real-time tasks.The proposed algorithm tried to lower the frequency of real-time tasks of higher priority by analyzing all the slack times.Experimental results demonstrate that this algorithm can save up to 26.2% more energy than the DPM algorithm.
Dynamic Reallocation Problems in Scheduling

Wei Quan Lim,Seth Gilbert,Wei Zhong Lim

DOI: https://doi.org/10.48550/arXiv.1507.01981

2016-02-16

Abstract:In this paper we look at the problem of scheduling tasks on a single-processor system, where each task requires unit time and must be scheduled within a certain time window, and each task can be added to or removed from the system at any time. On each operation, the system is allowed to reschedule any tasks, but the goal is to minimize the number of rescheduled tasks. Our main result is an allocator that maintains a valid schedule for all tasks in the system if their time windows have constant size and reschedules O(1/{\epsilon}*log(1/{\epsilon})) tasks on each insertion as {\epsilon}->0, where {\epsilon} is a certain measure of the schedule flexibility of the system. We also show that it is optimal for any allocator that works on arbitrary instances. We also briefly mention a few variants of the problem, such as if the tasks have time windows of difference sizes, for which we have an allocator that we conjecture reschedules only 1 task on each insertion if the schedule flexibility remains above a certain threshold.

Data Structures and Algorithms

Real Time Performance Evaluation for Dynamic Scheduling Algorithm

An Efficient Fault-Tolerant Scheduling Algorithm for Periodic Real-Time Tasks in Heterogeneous Platforms

A Reliability Model for RealTime Multitasking Software

Research on Real-Time Scheduling Strategy for Transient Fault Tolerance in Nc System

Control Oriented Direct Feedback Scheduling

Dynamic Task Scheduler for Real Time Requirement in Cloud Computing System.

Real-time scheduling for tasks with uncertain execution times

Research on Dynamic Time Evaluation of Service Procedure and Task Scheduling

Survey of Weakly-Hard Real Time Schedule Theory and Its Application

A Decomposition-Based Approach for Multitask Scheduling with Execution Uncertainty in Industrial Internet of Things

Efficient Scheduling Algorithm for Hard Real-Time Tasks in Primary-Backup Based Multiprocessor Systems

Time Analysis for Fault-Tolerant Real-Time Group-Based Scheduling

A Real-Time Scheduling Algorithm for Embedded Systems with Various Resource Requirements

Partial preemptive real-time scheduling algorithm in software fault-tolerant model

Study and Implementation of Real-Time Scheduling Mechanism Based on Linux Kernel

Fault-tolerant real-time tasks scheduling with dynamic fault handling

Reliable Estimation of Execution Time of MapReduce Program

Research on Dynamic Feedback and Elastic Scheduling Model and Algorithm for Flexible Workload

Performance evaluation of a dynamically adaptive co-scheduling algorithm

A New Energy-Efficient Online Scheduling Algorithm for Static Priority Real-Time Systems

Dynamic Reallocation Problems in Scheduling