Fault-tolerant Gang Scheduling in Distributed Real-time Systems Utilizing Imprecise Computations

Distributed real-time systems play an increasingly vital role in our society. The most important aspect of such systems is the scheduling algorithm, which must guarantee that every job in the system will meet its deadline, providing high-quality (precise) results. In this paper we evaluate by simulation the performance of strategies for the scheduling of parallel jobs (gangs) in a homogeneous distributed real-time system with possible software faults. For each scheduling policy we provide an alternative version which allows imprecise computations. We propose a performance metric applicable to our problem, which takes into account the number of jobs guaranteed, as well as the precision of the results of each guaranteed job. The simulation results show that the alternative versions of the algorithms outperform their respective counterparts. To the best of our knowledge, a real-time gang scheduling approach that utilizes imprecise computations has never been discussed in the literature before.