Two-stage single-instance data de-duplication backup method

The invention discloses a two-stage single-instance data de-duplication backup method. De-duplication of data at two stages is performed during backup. The method comprises firstly, performing repeated data detection on a file scale, inquiring local logs to judge whether identical files are stored or not, and if the identical files are stored, informing users to complete the backup operation; if identical files are not stored locally, informing backup programs of server ends to inquire databases to judge whether files with identical content exist or not, if the files with identical content are searched, establishing links pointing to the files for clients only, and recording quotes of the files by clients by the server ends; if the files are new, uploading the files and recording information of the files by two ends; further processing the files after the files are uploaded to the server ends by background programs, and splicing small files together to avoid waste of space; storing large files respectively by type, comparing similar files regularly, and performing difference de-duplication at the second stage after grouping.