Data Restoration via Deduplication Index Pre-Query
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data restoration methods from deduplicated storage frequently access storage devices, leading to increased communication load, reduced efficiency, and shortened storage device lifespan due to repeated searches for deduplication indices during data restoration.
Innovation Solution
A data restoration method where the backup server searches for metadata and deduplication indices, determines the corresponding data slices, and writes them to a target disk based on the deduplication index and metadata, reducing redundant access to the storage device.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the backup server sequentially searches for the deduplication index in a storage order of all data slices, then the backup server can ensure data restoration correctness, but the storage device is frequently accessed which increases communication load and reduces restoration efficiency
Solution Approach 1:
The patent applies preliminary action by pre-obtaining all deduplication indices corresponding to metadata of data slices before the actual data restoration process. The backup server queries the storage device once to obtain multiple deduplication indices in advance, stores them locally, and then uses these pre-obtained indices for restoration without needing to repeatedly query the storage device. This resolves the contradiction by maintaining restoration correctness through complete index acquisition while improving efficiency by eliminating repeated access operations.
2Reliability
If the backup server reads data slices multiple times from the storage device during restoration, then all data slices can be restored, but the service life of the storage device is shortened due to frequent access
Solution Approach 1:
The patent applies preliminary action by obtaining all necessary deduplication indices before the restoration process begins. The backup server queries the storage device once to retrieve all deduplication indices corresponding to the metadata, stores them in memory, and then performs restoration using these pre-obtained indices. This single preliminary query ensures complete data restoration while minimizing storage device access frequency, thereby extending storage device service life.
3Reliability
If the backup server performs frequent communication with the storage device during data restoration, then data slices can be located and restored, but the communication load increases and restoration efficiency decreases
Solution Approach 1:
The patent applies preliminary action by performing a single preliminary communication operation to obtain all deduplication indices before restoration. The backup server queries the storage device once with the backup data identifier to retrieve all deduplication indices corresponding to the metadata of data slices, stores these indices locally, and then performs restoration operations using the pre-obtained indices without further communication with the storage device. This approach ensures accurate data slice location while dramatically reducing communication load and improving restoration efficiency.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Embodiments of the present invention provide a data restoration method and apparatus. After a data restoration request is received, all metadata corresponding to each deduplication index is determined based on a deduplication index corresponding to metadata of each data slice in backup data. Then, after a data slice to which the deduplication index points and that is stored in a storage device is obtained, the data slice is stored to a distribution location described in each piece of metadata corresponding to the deduplication index. According to the embodiments of the present invention, for restoration of a plurality of data slices that have same data content and that correspond to a same deduplication index, the storage device is prevented from being frequently accessed to obtain, many times, the data slice to which the deduplication index points and that is stored in the storage device. In this way, a service life of the storage device is prolonged, load of communication with the storage device is reduced, and data restoration efficiency is improved.