Stabilizing Pre-Run-Time Schedules With the Help of Grace Time

This paper discusses the stability of a feasible pre-run-time schedule for a transient overload introduced by processes re-execution during an error recovery action. It shows that the stability of a schedule strictly tuned to meet hard deadlines is very small, invalidating thus backward error recovery. However, the stability of the schedule always increases when a real-time process is considered as having a nominal and a hard deadline separated by a non-zero grace time. This is true for sets of processes having arbitrary precedence and exclusion constraints, and executed on a single or multiprocessor based architecture. Grace time is not just the key element for the realistic estimation of the timing constraints of real-time error processing techniques. It also allows backward error recovery to be included in very efficient pre-run-time scheduled systems when the conditions stated in this paper are satisfied. This is a very important conclusion, as it shows that fault-tolerant hard real-time systems do not have to be extremely expensive and complex.     

一种有限优先级的静态优先级分配算法

静态优先级调度在实时系统中得到了广泛应用.然而,静态优先级调度受到系统支持的优先级个数的限制.当任务的个数大于优先级个数时,需要将多个任务映射到同一个优先级.针对优先级个数有限的情况,给出了在截止期限大于周期时任务可调度的充分必要条件,并提出了基于有限优先级的静态优先级分配算法(AGP).AGP算法对于基本任务集合是最优的静态优先级分配算法.其最优性表现在,所需的优先级个数最小,并且若采用AGP算法不可调度某个任务集,则采用其他静态优先级分配算法也不可调度该任务集.模拟结果表明,AGP算法的可调度性要远远大于常量法.AGP算法对于解决在嵌入式实时系统中任务的优先级分配问题具有重要意义.     

容错优先级可提升的抢占阈值容错调度算法

基于软件容错模型,提出了允许容错优先级提升的抢占阈值容错调度算法(extended fault-tolerantfixed-priority with preemption threshold,简称FT-FPPT*).该算法能够在抢占式容错调度算法(fault-tolerantfixed-priority preemptive,简称FT-FPP)和抢占阈值容错调度算法(fault-tolerant fixed-priority with preemptionthreshold,简称FT-FPPT)无法提高系统容错能力的情况下,进一步提高系统的容错能力.为了获得系统中任务优先级分配的最佳策略,基于任务最坏响应时间的可调度性分析,提出了一种最优的优先级配置搜索算法(priorityassignment search algorithm,简称PASA).经过深入分析和实验证明,与FT-FPPT算法相比,FT-FPPT*算法能够有效地提高硬实时系统的容错能力.     

副版本优先级可提升的全局容错调度算法

在主副版本机制的全局容错调度中,副版本运行窗口短,采用优先级继承策略的副版本响应时间长,容易错失截止期.针对副版本实时性差的问题,提出基于优先级提升策略的全局容错调度算法(fault tolerant global scheduling with backup priority promotion, FTGS-BPP),通过赋予副版本比主版本高的优先级,减少副版本在运行过程中受到的干扰,缩短了副版本的响应时间,改善了副版本的实时性,从而减少了实现容错所需的额外处理器资源.仿真结果表明,和采用优先级继承策略的全局容错调度算法相比,FTGS-BPP在调度相同的任务集时明显降低了处理器资源需求.     

一种静态最少优先级分配算法

随着实时系统越来越多地应用于各种快速更新系统,尤其是各种片上系统,如PDA(personal digital assistant),PSP(play station portable)等,性价比已成为系统设计者的主要关注点.实际应用中,实时系统通常仅支持较少的优先级,常出现系统优先级数小于任务数的情况(称为有限优先级),此时,需将多个任务分配到同一系统优先级,RM(rate monotonic),DM(deadline monotonic)等静态优先级分配算法不再适用.为此,静态有限优先级分配是研究在任务集合静态优先级可调度的情况下,可否以及如何用较少或最少的系统优先级保持任务集合可调度.已有静态有限优先级分配可分为两类:固定数目优先级分配和最少优先级分配.给出了任意截止期模型下任务静态有限优先级可调度的充要条件以及不同静态有限优先级分配间转换时的几个重要性质,指出了系统优先级从低到高分配策略的优越性,定义了饱和任务组与饱和分配的概念,证明了在任务集合静态优先级可调度的情况下,最少优先级分配比固定数目优先级分配更具一般性.最后提出一种最少优先级分配算法LNPA(least-number priority assignment).与现有算法相比,LNPA适用范围更广,且复杂度较低.     

具有双优先级的信元调度系统中的费用优化问题

本文对具有双优先级的信元调度系统的费用问题进行了量化分析,得到了重要结论：某种策略使得系统丢包最少但不一定总是使系统收益最大,并提出了具体的得到最优服务水平的优化解决方案。     

基于时延的动态优先级调度算法

队列管理是提高网络QoS的一种有效方法。在基于时延的调度算法（BDS）基础上将时间片与优先级相结合,提出了一种基于时延的动态优先级调度算法（DDPQS）。为了实现该算法,针对进入缓冲区的每个子队列设置一个计数器,以调整的计数器值为基准来动态的改变队列的优先级,从而达到队列调度的效果;又从研究该算法的过程中,发现其局限性,即计数器值对时间片过于敏感的问题,于是进一步采用设置阈值进行区分的方法来优化。优化前后的仿真结果表明,时延和吞吐率性能具有明显改善。     

Priority inheritance in soft real-time databases

Due to resource sharing among tasks, priority inversion can occur during priority-driven preemptive scheduling. In this work, we investigate solutions to the priority inversion problem in a soft real-time database environment where two-phhse locking is employed for concurrency control. We examine two basic schemes for addressing the priority inversion problem, one based on priority inheritance and the other based on priority abort. We also study a new scheme, called conditional priority inheritance, which attempts to capitalize on the advantages of each of the two basic schemes. In contrast with previous results obtained in real-time operating systems, our performance studies, conducted on an actual real-time database testbed, indicate that the basic priority inheritance protocol is inappropriate for solving the priority inversion problem in real-time database systems. We identify the reasons for this performance. We also show that the conditional priority inheritance scheme and the priority abort scheme perform well for a wide range of system workloads.This work was supported by the National Science Foundation under Grant IRI-8908693 and by the U.S. Office of Naval Research under Grant N00014-85-K0398.A previous version of this paper appeared in Real-Time Systems Symposium, Dec. 1991.     

高速信元优先级交换调度策略定性研究

该文根据流平衡定律和排队理论用近似求解的方法对带优先级的信元调度策略进行了定性分析研究,推导得出了更加符合实际的解析式,并通过仿真和数值计算进行了验证。     

基于优先级的访存调度算法研究

文章研究了存储控制器中的访存调度策略,提出了基于优先级的访存调度算法。首先使用遗传算法建立有效的数据源,然后对得到数据源应用统计进行调度优先级挖掘,共获取三个优先级别,这样仅使用这三个优先级构造调度算法进行访存序列调度。实验结果表明,提出的算法很好地降低了访存序列的运行时间,优化效果接近于文献[4]中提出的贪婪访存调度算法,但算法运行时间却远小于后者。