Proactive Software Rejuvenation Solution for Software Fault Tolerance System

Abstract

Software failures are now known to be a dominant source of system outages. The cost of unplanned outages is higher than the cost of planned outages. Software systems often experience aging related defects which cause unexpected outages and affecting its availability. A system should have the capability enough to tolerate the faults and operate continuously in spite of faults and interruptions. A proactive fault management method to deal with the software aging phenomenon is software rejuvenation. In this paper, we focus on a high assurance software system with fault-tolerance and proactive software rejuvenation and analyze the stochastic behavior of such a software system. More precisely, we consider a software fault tolerance system with redundant and evaluate the steady-state system availability based on the familiar Markovian analysis through the use of numerical analysis.