A scalable on-line multilevel distributed network fault detection/monitoring system based on the SNMP protocol

Traditional centralized network management solutions do not scale to presentday large-de computerlcommunleation networks. They suffer from certain other drawbacks tw: a *@e point of failure and hence lack of fault toleranee, and heavy eommunieadon costs asraeiated with the central manager. It har been recognized that decentrallzatloddlstributed solutions can solve some of the problem arroeiated with centralized solutions [3]. For this reason, there has been considerable interest in the recent past on distlibutedidecentrdized network management applications. Our work in this paper has been motivated by this research trend In network management. We present the design and evalvation of an SNMP-had didliboted network fault detectiodmonltorlng system. The design involver the integration of our reeentiy developed ML-ADSD algorithm 1111, 1161 for diagnosis of faults In a distributed system of processors into the SNMP framework The ML-ADSD algorithm uses the multilevel pandigm and Is scalable in the sense that only minor modifications wlU be required to adapt the algorithm to networlu of varying sizes. The system allows processors to fail andlor recover during the process ofdiagnosir. Thus the system has fault tderance capablllty. We demonstrate the application of the system by Implementing It on an Ethernet network of 32 maehinei. Our results ertsblish that the diagnosis latency (or tlme to termlmtion) is mueh bener than the latency of e d e r solutions. Also, the bandwidth utilization of our system Is very insignificant, thereby demonstratlng the practicality of deployment ofthe system in a real network environment. Thus in this work we have successfully Integrated three modern disciplines: network management, dishihoted computing and system level diagnosis.