Thin Chunk Signature Lists for Backup Object Size Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The inefficiency in generating and managing backups for fully provisioned virtual volumes, where significant storage space is allocated but much is unused, leading to poor performance in complete and synthetic complete backup operations due to the inclusion of unused data patterns.
Innovation Solution
The generation of thin lists of chunk signatures by omitting signatures for unused regions and creating usage metadata, allowing for more efficient backup operations by storing only used data regions, thereby reducing the size of backup objects and improving performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If complete backup operations include all allocated storage space for fully provisioned volumes, then data protection coverage is ensured, but backup operation time and processing resources increase significantly due to unused data patterns
Solution Approach 1:
The patent extracts and removes unused data patterns from the backup process by comparing chunk signatures against a database of known unused patterns. Only chunks representing actually used data are included in backup operations, while unused chunks are excluded based on signature matching, thereby reducing backup time without compromising protection of actual data.
Solution Approach 2:
The patent applies different treatment to different regions of the storage volume based on their usage status. Used regions are backed up with full protection, while unused regions are identified and excluded from backup operations. This local differentiation allows the system to maintain data protection for actual data while avoiding unnecessary processing of unused space.
2Reliability
If backup objects store all data chunks from fully provisioned volumes, then complete data restoration is enabled, but backup object size becomes excessively large due to unused regions
Solution Approach 1:
The patent extracts unused data patterns from backup objects by comparing chunk signatures against a database of unused pattern signatures. Chunks matching unused patterns are excluded from backup objects, significantly reducing backup object size while maintaining the ability to restore all actually used data through the remaining chunks.
Solution Approach 2:
The patent uses signature-based copying where instead of storing actual unused data chunks, the system creates references or metadata indicating which chunks are unused. This allows the backup system to represent unused regions without consuming storage space for actual chunk data, reducing backup object size while preserving restoration capability for used data.
3Reliability
If synthetic complete backup operations process all volume capacity, then accurate backup state is achieved, but processing performance degrades due to unused data patterns
Solution Approach 1:
The patent extracts and removes unused data patterns from synthetic complete backup operations by comparing chunk signatures against the unused patterns database. This exclusion reduces the amount of data processing required for synthetic backups, improving processing performance while maintaining accurate representation of the actual backup state through the remaining used chunks.
Data Source
AI summary
Examples may include backup objects for fully provisioned volumes with thin lists of chunk signatures. Examples may generate one or more full lists of chunk signatures for the address space of a fully provisioned volume, compare each chunk signature of the full list to an unused region chunk signature representing a chunk of an unused region of the fully provisioned volume, generate metadata to indicate used regions of the fully provisioned volume, based on the comparisons, and generate from the one or more full lists, one or more thin lists omitting all chunk signatures determined to match the unused region chunk signature.