A method of localising a fault in a network is disclosed. The network comprises nodes (SW1 . . . SW4) links, and edge-nodes (EN1 . . . EN4) arranged as a plurality of spanning trees (T1, T2,T3), the spanning trees being partially disjoint. The network further comprises means for network management. The method comprising the steps of receiving information on the configuration of the plurality of tree topologies in the network; monitoring connectivity in the network; upon detection of a loss of connectivity in the network, identifying the failed tree(s), and determining the network elements common to the failed tree(s).