论文部分内容阅读
The problem of adapting backward error recovery to parallel real time systems is discussed in thispaper. Because of error propagation among different cooperating processes, an error occurring in oneprocess may influence some important outputs in other processes. Therefore, a local output has to bedelayed until its validity is confirmed globally. Since backward error recovery adopts redundancy ofcomputing time instead of processing equipment, the variation of the actual execution time of acooperating process may be very large if it works in an unreliable environment. These problems are theprimary obstacles to be removed. Previous studies focus their attentions on how to eliminate domino-effectdynamically. But backward error recovery cannot be applied directly in parallel real time systems evenunder the condition that no domino-effect exists. How to reduce output delays efficiently if nodomino-effect remains? How to estimate this delay time? How to calculate the actual execution timeof every process and how to schedule these processes under an unstable condition? These problems wereomitted in literature unfortunately. The interest of this paper is to provide satisfactory solutions to theseproblems to make it possible to adopt backward error recovery efficiently in parallel real time systems.
The problem of adapting backward error recovery to parallel real time systems is discussed in this paper. Since of error propagation among different cooperating processes, an error output in differentprocrating processes, an error validity is yet taking globally. Since backward error recovery employs redundancy of computing time rather than processing equipment, the variation of the actual execution time of acooperating process may be very large if it works in an unreliable environment. These problems are the primary obstacles to be removed. Previous studies focus their attentions on how to eliminate domino-effectdynamically. But backward error recovery can not be applied directly in parallel real time systems evenunder the condition that no domino-effect exists. How to reduce output delays efficiently if nodomino-effect remains? How to estimate this delay time? How to calculate the actual execution time of e very process and how to schedule these processes under an unstable condition? These problems wereomitted in literature unfortunately. The interest of this paper is to provide satisfactory solutions to theseproblems to make it possible to adopt backward error recovery efficiently in parallel real time systems.