Flash Storage Erase Block Mapping for Cross-Size Rebuilding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge arises when replacing a flash storage device with a different erase block size, leading to issues with data representation and wastage of physical memory space due to mismatched block sizes.
Innovation Solution
A method for rebuilding data by determining and allocating erase blocks with differing sizes, mapping logical addresses, and rebuilding data into replacement blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a failed drive with one erase block size (e.g., 64 MB) is replaced with a new drive that has a different erase block size (e.g., 24 MB), then the system can continue operating with updated hardware, but data representation issues arise where replacement blocks start with data from the middle of the failed block, causing metadata corruption
Solution Approach 1:
The patent segments the data reconstruction process by identifying and handling the first replacement erase block separately from subsequent replacement blocks. The first block is allocated to start at a specific boundary to ensure proper metadata alignment, while subsequent blocks are allocated sequentially. This segmentation prevents metadata corruption by ensuring that each replacement block starts with properly aligned data boundaries.
Solution Approach 2:
The patent introduces an intermediary mapping mechanism that translates logical addresses to physical replacement block locations. This mapping layer acts as a mediator between the logical data structure and the physical replacement blocks, ensuring that data is correctly positioned and metadata is properly aligned even when replacement blocks have different sizes than the original failed block.
2Reliability
If replacement blocks are allocated to match the capacity of the original failed block, then data can be rebuilt accurately, but physical memory space is wasted when the replacement set has more capacity than the original failed block
Solution Approach 1:
The patent changes the parameter of block allocation from fixed-size matching to flexible size adaptation. Replacement blocks are allocated with sizes matching the new drive's erase block size rather than forcing them to match the original failed block size. The system adjusts the allocation parameters to utilize the full capacity of the replacement drive, eliminating physical memory wastage while maintaining data reconstruction accuracy through proper address mapping.
Solution Approach 2:
The patent introduces a logical address mapping dimension that decouples the relationship between physical block sizes and logical data organization. By adding this addressing layer, the system can accommodate replacement blocks of different sizes without compromising data integrity, allowing full utilization of the replacement drive's capacity while maintaining accurate data reconstruction.
3Productivity
If replacement erase blocks are allocated sequentially without considering boundary alignment, then allocation is simpler and faster, but systems that represent liveness of a block with metadata written at the start of the block experience data representation issues
Solution Approach 1:
The patent performs preliminary action by pre-calculating and recording the allocation boundaries of replacement erase blocks during the allocation phase. The system identifies which replacement blocks will serve as the first block for each original block and stores this boundary information in the mapping structure. This preliminary preparation ensures that when data is rebuilt, the boundary alignments are already established, maintaining both allocation efficiency and data alignment precision without requiring complex real-time calculations.
Data Source
AI summary
A method for rebuilding data when changing erase block sizes in a storage system is provided. The method includes determining one or more erase blocks to be rebuilt and allocating one or more replacement erase blocks, wherein the one or more erase blocks and the one or more replacement erase blocks have differing erase block sizes. The method includes mapping logical addresses, for the one or more erase blocks, to the one or more replacement erase blocks and rebuilding the one or more erase blocks into the one or more replacement erase blocks, in accordance with the mapping.