Abstract: A backup storage system includes a server having a restore operation that is independent from browse, mark and submit operations. A restore engine process is created as part of a restore session initiated by a client for retrieving a list of restorable objects to be displayed to a user via a graphical user interface. The user can browse the list of restorable objects, mark or select objects for restoration and submit the list of mark objects. Multiple restore submissions can be made prior to the user commanding the restore engine process, via a remote procedure call, to execute the submitted restore requests. To support new backup techniques and storage devices, a corresponding library can be added for processing with minimal impact to the overall restore system.

Abstract: A backup and restore system includes a work item server for restoring multiple work times simultaneously. The work item server assigns each work item to a respective trail queue based upon the commonality of the volumes on which the work items are stored. The work item server allocates drives to the trail queues for restoring multiple work items simultaneously. The trail queue have drives assigned for work items in that trail queues to use during a restore. The trail queue has a concurrency defined to predetermine the number of simultaneous work items can run that are associated with the trail queues. Drives can be allocated to the trail queues for executing work item restores associated with the respective trail queues until a predetermined trail queue concurrency is reached.

Abstract: A backup and restore system forms a persistent submit object from objects marked for restoration by a user. The submit object encapsulates the information needed to restore the work items contained in the submit object. The submit object enables restore execution to be independent from the browse, mark and submit operations. The submit object also allows multiple restore requests to be submitted.