论文部分内容阅读
在硬实时系统的应用中 ,如果硬实时任务不能在规定的时限完成 ,将会产生人员伤亡 ,财产损失等严重后果 ,为了保证在系统出错的情况下 ,硬实时任务仍然能在截止时限之前完成 ,必须研究实时容错技术 .本文从实时容错调度算法的角度出发 ,提出一种基于分布式系统的实时容错调度算法 ,并研究了该算法的时间复杂度 ,同时给出一个实例说明该容错调度算法的调度过程 .这种容错调度算法称为“无容错需求后调度算法”( NFRL ) ,该实时容错调度算法可以同时调度具有容错需求的实时任务和无容错需求的实时任务 .由算法 NFRL 所产生的调度可保证在分布式系统中一个站点机失效的情况下 ,具有容错需求的实时任务仍然可在截止时间内完成 .
In the application of hard real-time system, if the hard real-time tasks can not be completed within the prescribed time limit, there will be serious consequences of personal injury and property loss. In order to ensure that the hard real-time tasks can still be completed before the deadline in the event of a system error , We must study the real-time fault-tolerant technology.In this paper, a real-time fault-tolerant scheduling algorithm based on distributed system is proposed from the perspective of real-time fault-tolerant scheduling algorithm, and the time complexity of the algorithm is studied.An example is also given to illustrate the fault tolerant scheduling algorithm This fault tolerant scheduling algorithm, called “Post-fault-free scheduling algorithm” (NFRL), real-time fault-tolerant scheduling algorithm can simultaneously schedule real-time tasks with fault-tolerant requirements and real-time tasks without fault tolerance. Scheduling ensures that real-time tasks with fault-tolerant requirements can still be completed by the deadline in the event of a site machine failure in a distributed system.