Fault management for Internet Services: Modeling and Algorithms

A modeling approach is proposed in this paper to build the bipartite fault propagation model (FPM) for Internet services. The FPM is layered as Internet services involve multiple layers. Two fault localization algorithms, MCA (Max-Covering Algorithm) and MCA+, are designed for the bipartite FPM. MCA+ is an extension of MCA, taking lost and spurious symptoms into account. Simulation results show that MCA+ achieves high detection rate, low false positive rate and has polynomial computational complexity even in the presence of lost and spurious symptoms.