Shallow Copy via Global Region for Cross-Partition Metadata
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current copy mechanisms, such as deep object copies, are inefficient and wasteful of cluster throughput, especially when copying large objects, and can result in unnecessary storage consumption and dangling pointers during shallow copies across bucket partitions.
Innovation Solution
Implementing a shallow copy method that copies only object metadata, creating a global region that points to the source region and allowing the destination region to reference the global region, thereby avoiding data deletion and dangling pointers, and enabling cross-partition metadata separation without expensive data copies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If deep object copy is used, then data integrity is maintained, but storage consumption increases and cluster throughput decreases
Solution Approach 1:
The patent implements shallow copying of object metadata instead of deep copying of object data. The metadata service creates a copy of the object's metadata (including region information, bucket identifier, partition identifier, and region identifier) without copying the actual object data, thereby maintaining data integrity references while reducing storage consumption and improving cluster throughput.
2Productivity
If shallow copy is used across bucket partitions, then copy efficiency improves, but dangling pointers may occur
Solution Approach 1:
The patent introduces a global region as an intermediary structure that mediates between source and destination bucket partitions. The global region contains a reference to the source region and is stored in the destination partition, allowing the destination object to reference the source region through the global region without creating dangling pointers. This intermediary mechanism enables efficient cross-partition copying while maintaining pointer integrity.
Solution Approach 2:
The patent adds a global region dimension to the traditional bucket-partition structure. Instead of directly mapping objects within a single partition, the system introduces a global region layer that spans multiple partitions. This dimensional change allows metadata to be separated and managed across partitions while maintaining consistent references through the global region, preventing dangling pointers during shallow copies.
3Quantity of substance
If shallow copy is used, then storage consumption reduces, but data integrity may be compromised
Solution Approach 1:
The patent copies only the essential metadata fields (region information, bucket identifier, partition identifier, region identifier) rather than the entire object data. This selective copying approach reduces storage consumption significantly while maintaining data integrity through proper reference management via the global region structure.
Data Source
AI summary
An illustrative embodiment disclosed herein is an apparatus including a processor having programmed instructions to create a global region that is associated with a first bucket partition and a second bucket partition different from the first bucket partition, provide, to the global region, region information of a source region in which a source object is stored, create a destination region in which a destination object is stored, and provide, to the destination region, from the global region, the region information of the source region. In some embodiments, the source region is in the first bucket partition and the destination region is in the second bucket partition.