Abstract: A method in a processing system for managing backup data on a set of nodes in a network data processing system. An inventory of files is generated on a set of nodes in the network data processing system. An initial map from the inventory is created, wherein the first initial map includes an identification of each file located on a node in the set of nodes. A location map is built from the initial map and the location map identifies a set of files and identifies each node on which a file in the set of files is located, wherein at least one node in the set of nodes is a peer node to another node in the set of nodes.