An eminent approach of fault management using proactive and reactive techniques in distributed computing

Abstract

Many of the existing distributed systems perform remote monitoring, and even generate alarms when fault occurs. But they fail in finding the exact location of the fault, or even the automatic execution of appropriate fault recovery actions. Hence for large distributed systems with many computing nodes it may be rather time-consuming and difficult to resolve the problems in a short time by an exhaustive search in order to find the cause for failure. At the time of failure many problems like loss or delay of fault messages may occur. Also a failure may result in a number of unreliable alarms. A good sophisticated design for fault management should work out efficiently whenever there is a redundant and incomplete data. So we proposed a system comprising of several components where we have a fault detection engine in which various techniques have been proposed for detecting faults in distributed computing. Once the faults are detected, one may diagnose the system to track the root cause. For diagnosis the system we are going to implement that In Expert system and we are combining both the reactive and proactive techniques for fault management. © 2013 Springer-Verlag.