Parallel file restoration from a serialized storage device using a network data management protocol

A method, comprising:

• copying, by a computer system comprising a processor, data representing respective files that have been stored in a tape drive from the tape drive to memory buffers as buffered data;
• determining, by the computer system, file header records representing locations of the respective files in the memory buffers;
• based on the file header records, generating and storing, by the computer system, respective work items in a work item queue for facilitating restoration of the respective files;
• and
• independently recovering, utilizing multiple writers of a group of writers of the computer system, multiple files of the respective files to the computer system, wherein the independently recovering comprises
• storing a header work item of the respective work items in the work item queue, retrieving the header work item from the work item queue, and
• creating, based on a file header record of the file header records, a file object data structure comprising information for facilitating restoration of a file of the respective files, wherein the header work item of the respective work items comprises an address pointer for locating the file object data structure within a file list, and wherein the multiple writers independently retrieve, in order of tape offsets of the multiple files within the tape drive, the respective work items from the work item queue to facilitate parallel recovery, by the multiple writers, of the multiple files from the memory buffers according to the order of the tape offsets of the multiple files within the tape drive.

A non-transitory computer-readable storage device having stored thereon executable instructions that, in response to execution, cause a device comprising a processor to perform operations, the operations comprising:

• copying data representing respective files that have been stored in a tape drive from the tape drive to memory buffers as buffered data;
• determining file header records representing locations of the respective files in the memory buffers;
• based on the file header records, generating and storing respective work items in a work item queue for facilitating restoration of the respective files;
• and
• independently recovering, utilizing multiple writers of a group of writers of the device, multiple files of the respective files to the device, wherein the independently recovering comprises
• storing a header work item of the respective work items in the work item queue,
• retrieving the header work item from the work item queue, and
• creating, based on a file header record of the file header records, a file object data structure comprising information for facilitating restoration of a file of the respective files, wherein the header work item of the respective work items comprises an address pointer for locating the file object data structure within a file list, and wherein the multiple writers independently retrieve, in order of tape offsets of the multiple files within the tape drive, the respective work items from the work item queue to facilitate parallel recovery, by the multiple writers, of the multiple files from the memory buffers according to the order of the tape offsets of the multiple files within the tape drive.