SAN-to-NAS Data Migration Using Metadata-Based Virtual Copy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data migration methods between storage systems, especially within the same storage array, require unnecessary data copying, leading to inefficient use of processing resources and bandwidth, and can create duplicate data copies, wasting storage space.
Innovation Solution
Utilizing virtual copy logic that associates data blocks from a block-based storage system with a file-based system by updating metadata, allowing data migration without physical data movement, and optionally renumbering blocks for efficient metadata layouts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional data copying methods are used to migrate data between storage systems, then data integrity is preserved, but processing resources and bandwidth are wasted due to unnecessary data copying
Solution Approach 1:
The patent applies virtual copying by creating metadata references instead of physical data copies. The virtual copy logic generates metadata that points to the original data blocks, allowing the storage system to access data from the source location while appearing to copy it to the destination. This eliminates actual data movement while maintaining data integrity through reference validation.
Solution Approach 2:
The patent introduces virtual copy logic as an intermediary component between the source and destination storage systems. This intermediary creates and manages metadata that mediates access to the original data, allowing both systems to reference the same physical data blocks without requiring actual data transfer, thus preserving integrity while improving efficiency.
2Reliability
If conventional data copying methods are used to migrate data between storage systems, then data availability is maintained, but bandwidth is consumed due to physical data movement
Solution Approach 1:
The patent implements virtual copying where metadata is created to reference original data blocks rather than physically copying the data. The virtual copy logic generates metadata entries that point to the source storage locations, allowing the destination system to access data without actual bandwidth consumption while maintaining data availability through the reference mechanism.
Solution Approach 2:
The patent extracts only the essential metadata from the data migration process, separating the metadata creation function from the physical data copying function. By taking out just the metadata generation and reference management components, the system achieves data availability without the bandwidth overhead of moving actual data blocks.
3Reliability
If conventional data copying methods are used to migrate data between storage systems, then complete data duplication is achieved, but storage space is wasted due to duplicate data copies
Solution Approach 1:
The patent applies virtual copying by creating metadata references that point to the original data blocks in the source storage system. Instead of duplicating physical data, the virtual copy logic generates metadata entries that reference the same data blocks, achieving complete data accessibility while using minimal storage space for the metadata structures themselves.
Solution Approach 2:
The patent implements a nested structure where metadata is embedded within the storage system's existing metadata frameworks. The virtual copy metadata is nested within the destination volume's metadata structures, allowing the system to maintain complete data references without external storage overhead, effectively nesting the copy management within the existing storage architecture.
Data Source
AI summary
A request is received from an initiator application to migrate data from a plurality of blocks of a storage area network (SAN) volume resident on a storage array to a network attached storage (NAS) volume resident on the storage array. Metadata is generated for the NAS volume, the metadata to reference the plurality of data blocks of the SAN volume storing the data. A file is generated at the NAS volume using the metadata, wherein the file is associated with the plurality of data blocks of the SAN volume storing the data.