CRC-Seeded Deduplication Checksums for Data and Stub Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In deduplication processes, distinguishing between actual data and stubs pointing to parent copies is crucial for accurate verification and performance, but existing methods lack effective mechanisms to differentiate between them efficiently.
Innovation Solution
The use of differing cyclical redundancy check (CRC) seeds, where one CRC seed is used for actual data and a numerically different CRC seed is used for stubs, allows for verification and differentiation between data and stubs by generating unique checksums.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of stationary object
If data deduplication is implemented using stubs pointing to parent copies, then storage capacity is reduced, but the ability to accurately verify and distinguish between actual data and stubs deteriorates
Solution Approach 1:
The patent applies local quality by assigning different CRC seed values to different types of data structures. Actual data blocks use one CRC seed value while stubs use a different CRC seed value. This local differentiation allows the verification process to distinguish between data and stubs by checking which seed value produces a matching checksum, thereby maintaining verification accuracy while preserving storage efficiency through deduplication.
2Productivity
If a large number of stubs are created in relation to actual data, then storage efficiency is improved, but the complexity of verifying data integrity increases
Solution Approach 1:
The patent changes the parameter of CRC seed values to differentiate between data and stubs. By using distinct seed values (e.g., first seed value for data, second seed value for stubs), the verification process becomes simpler despite the large number of stubs. The system can quickly determine whether to verify using one seed or the other based on the structure being examined, reducing the complexity burden that would otherwise arise from handling numerous stubs.
Data Source
AI summary
Various embodiments for differentiating between data and stubs pointing to a parent copy of deduplicated data. Undeduplicated data is stored with a checksum of an initial value. A stub pointing to the parent copy of the deduplicated data is stored with an additional checksum of a differing, additional initial value.