Abstract: The present invention provides a system and method for creating and accessing a copy of one or more host-accessible storage entities contained in a persistent consistency point image (PCPI) of a volume served by a storage system. The storage entity may comprise a file system, logical volume, volume group, or “file system on raw device”. The copy of the storage entity is generated using a logical unit number (lun) copying technique in conjunction with a PCPI table of contents (PTOC). The PTOC contains data describing contents of the PCPI, including luns of the storage entity. The PTOC is written to the volume of the storage system, which is configured to store the luns, i.e., in the context of an active file system, immediately prior to generation of the PCPI. After generation of the PCPI, the PTOC is deleted from the volume such that it is only stored within the PCPI and not within the active file system.

Abstract: A method and apparatus for operating a plurality of computers is disclosed. A file is maintained on a selected computer of the plurality of computers. An open request for the file is sent by a first computer of the plurality of computers to the selected computer, the open request including a first token. The file is opened for the first computer in response to the open request. An open request is sent for the file by a second computer to the selected computer, the open request including a second token. The first token is compared with the second token. If the first token and the second token match, opening the file for the second computer, and if the first token and the second token do not match, not opening the file for the second computer.

Abstract: A mechanism for controlled sharing of files by clustered applications is provided. The mechanism expands a distributed file access protocol, such as the direct access file system protocol, by including an open with share token command and an open_downgrade operation to adjust the access and deny levels for a given resource.

Abstract: A mechanism enables distributed atomic creation of client-private files using a distributed file system protocol. The mechanism provides an unlinked open operation that enables the distributed atomic creation of a client-private file that is not accessible by other clients. That is, the unlinked open operation creates a file, but does not generate a path name to the file within a file system. The unlinked open operation thus permits clients utilizing the file system protocol to create temporary or “scratch files” that are not accessible by other clients.