Abstract: One or more aspects of the disclosure relate to techniques for accessing a file stored in a clustered computing network, where the file is stored in the clustered computing network in a plurality of related volumes that stem from a particular (e.g., parent) volume. In the method, a request to retrieve the file from the particular volume is received at the client. The client then determines whether the file is stored in a client memory so as to be identified with the particular volume. If the file is not stored in the client memory so as to be identified with the particular volume, then the file is retrieved from a different but related volume.

Abstract: Embodiments of the present invention include creating a subsequent backup copy of a data set by storing a full backup copy of the data set to reflect a state of a data set including blocks of data, at a first point in time. At a second point in time, after the first point in time, changes to the data set are identified. Next, as part of the subsequent backup copy, only each block of the data set which has been changed or added since the first point in time is stored.