METHOD FOR MANAGING FAILURE IN A NETWORK OF NODES BASED ON A LOCAL STRATEGY

摘要

The invention relates to a fault management method in a network of nodes (21-63), comprising, for each node considered of all or part of the nodes (21-53) of the network performing the same calculation: first, a local backup step of the state of this node considered (21-53), at a storage medium (31-56) of this node considered (21-53), the link between this storage medium (31 -56) and this node considered (21-53) can be redirected from this storage medium (31-56) to another node (21-63), then, if the node considered is faulty (21, 42), a step of recovering the local backup of the state of this node considered (21, 42), by redirecting said link between the node considered (21, 42) and its storage medium (31, 45) so as to connect said medium storage (31, 45) to an operational node (23, 43) different from the considered node (21, 42), this operational node (23, 43) being already in the process of performing said cal cul, the local backups of these nodes considered (21-53), used for the recovery steps, are consistent with each other so as to correspond to the same state of this calculation, then, if at least one node considered is faulty (21 , 42), a step of returning said local backup of this node considered (21, 42) to a new additional node (61-63) added to said network on the occasion of the failure of this node considered (21, 42) .