Rapid data deletion method based on source end deduplication

Abstract

The invention provides a rapid data deletion method based on source end deduplication. The method comprises two steps of deleting a backup set and cleaning a medium. According to the method, the specific frequency of recording each specific block is no longer used for the reference frequency of the data block; a backup set object records a data file existing in a referenced data block, a deletionfunction uses a delayed deletion strategy and comprises two steps of deleting the backup set and cleaning a medium, deletion operation is simplified, deletion performance is improved, time for the data block to exist in a repeated deletion library is prolonged as much as possible, and resource waste caused by frequent backup and deletion is avoided.

Description

technical field [0001] The invention belongs to the technical field of data deduplication, and in particular relates to a fast data deletion method based on source-side deduplication. Background technique [0002] Backup devices are always filled with a lot of redundant data. In order to solve this problem and save more space, deduplication technology has naturally become the focus of attention. The use of deduplication technology can greatly reduce the amount of stored data, thereby freeing up more backup space, so that backup data can be kept on the disk for a longer period of time, and the source-side deduplication technology can also save a lot of bandwidth during backup. A backup device for data protection must have the basic functions of backup, restore and delete. [0003] The feature of the deduplication function is that only one data block is kept in the deduplication library storing data, and each data block is different and unique. The data backed up using the ...

AI Technical Summary

Problems solved by technology

This approach has the following disadvantages: First, when recording the number of indexes of each block, the granularity is relatively fine. Every time a backup task is executed, the database needs to be accessed regardless of whether the data block is a new block (for new blocks To insert a new fingerprint record, the number of references to an existing block needs to be updated), which will affect the performance of the backup, even if the space in the actual deduplication library may not be released after the deletion operation (if the number of references is not 0, it cannot be truly clean up the data blocks on the disk); secondly, when deleting the backup set, since it is necessary to traverse all the deduplicated blocks used in the backup set and adjust the fingerprint tables corresponding to all blocks, the deletion task will be time-consuming longer

Therefore, the existing deletion scheme has relatively low performance and is not suitable for scenarios where backup and deletion are frequent

Images