Thin Provisioning Mapping Layer for Storage Space Reclamation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage systems face inefficiencies due to the overhead of managing duplicate data, as conventional deduplication methods increase latency and overhead in write and read operations, particularly in primary data storage, where data blocks are not stored sequentially and require traversing multiple locations for access.
Innovation Solution
A system comprising a virtual logic unit (vLUN) layer, a thinly provisioned mapped logical unit (mapped LUN) layer, and a mapping layer that facilitates efficient input/output access by remapping logical chunk addresses to physical chunk addresses, allowing for data reduction and deallocation of unneeded chunks, thereby reducing storage needs and improving performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If conventional deduplication methods are used, then storage space is reduced by eliminating duplicate data, but write and read latency increases due to the overhead of managing duplicate data and non-sequential storage
Solution Approach 1:
The patent segments storage into multiple storage units (e.g., disk drives, SSDs) that can be independently managed. Each storage unit maintains its own sequential structure, allowing the system to divide and conquer the deduplication problem while preserving sequential access benefits at the unit level.
Solution Approach 2:
The patent introduces an intermediary mapping layer that sits between the logical view and physical storage. This mapping layer translates logical sequential addresses to physical locations across multiple storage units, eliminating the need for applications to directly handle complex deduplication structures while still enabling space reclamation.
2Quantity of substance
If data is stored in non-sequential locations for deduplication, then duplicate data is eliminated, but access performance deteriorates due to traversing multiple locations
Solution Approach 1:
By segmenting storage into multiple units and allowing each to maintain sequential storage locally, the patent preserves high access performance within each segment while achieving deduplication across segments through the mapping layer.
Solution Approach 2:
The patent adds a new dimension to storage organization by introducing the mapping layer as an intermediate abstraction. This additional layer enables the system to achieve deduplication without sacrificing sequential access performance, as the mapping handles the complexity rather than the storage layout itself.
3Productivity
If thin provisioning is implemented, then storage allocation efficiency is improved, but the complexity of managing storage extents and mappings increases
Solution Approach 1:
The mapping layer acts as an intermediary that automatically manages the complexity of thin provisioning, extents, and mappings. This intermediary handles the complex bookkeeping and translation tasks, making thin provisioning transparent to both applications and storage hardware while maintaining allocation efficiency.
Data Source
AI summary
The invention provides a system to reclaim space identified as no longer in use and comprises a vLUN, a thinly provisioned mapped LUN, a mapping layer, and a data reduction engine. Chunks of data are stored at logical chunk addresses (LCAs) in the vLUN and are mapped to corresponding physical chunk addresses (PCAs) in the thinly provisioned mapped LUN. The data reduction engine performs a data reduction process on a first logical chunk of data stored at a first LCA in the vLUN, where the first logical chunk has a size that is a nonzero integer multiple of the size of the storage extent of the thinly provisioned mapped LUN. After the data reduction process, the PCA associated with the first logical chunk is no longer needed, and the thinly provisioned mapped LUN is instructed to deallocate the PCA associated with the first logical chunk that is no longer needed.