Deduplication

摘要

A method, system, and computer program product comprising determining if a file is greater than a predetermined size, based on a positive determination that the file is greater than the predetermined size, searching for an associated signature file for the file, and if the associated signature file is found for the file, perfecting some or all the signatures from the associated signature for the file to the cache.

主权项

1. A computer-implemented method comprising:
determining if a requested file is greater than a predetermined size;
based on a positive determination that the requested file is greater than the predetermined size, searching for an associated signature file for the file in a metadata file; wherein the metadata file is enabled to store signatures for a plurality of files; wherein the signatures correspond to chunks of the files; wherein for signature files stored in the metadata file, the metadata file records when a chuck of a file has the same signature as another chunk of another file; wherein the metadata file further enables the server cache to answer a request for an read IO to a chunk of a first file with a chunk a second file when the metadata file indicates the signatures of the chunks are the same; and
if the associated signature file is found for the file, prefetching some or all the signatures from the associated signature for the file into a server cache.