Metadata Protected Sparse Block Set for SSD Cache
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data management systems face challenges in efficiently updating and maintaining cache data across multiple devices while ensuring redundancy and minimizing input-output operations, particularly when cache devices become unavailable or fail.
Innovation Solution
A data management device with a cache system that implements a block set structure, where a header and meta-data are mirrored across cache devices for redundancy, while blocks are not mirrored, allowing the cache to operate as long as one device is operational, and updates are processed in batches to reduce IO operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If header and meta-data are mirrored across all cache devices, then reliability is improved, but device complexity increases
Solution Approach 1:
The cache data structure is segmented into three distinct components: header (containing block set information), meta-data (containing block allocation information), and actual data blocks. This segmentation allows selective mirroring of only the header and meta-data across cache devices, while data blocks are distributed without mirroring, thus improving reliability of control information without proportionally increasing overall system complexity
Solution Approach 2:
Copies of the header and meta-data are created and stored on each cache device in the block set, while actual data blocks are not copied. This selective copying approach ensures that essential control information is redundant and available even if one cache device fails, improving reliability without requiring full data duplication across all devices
2Stability of the object's composition
If updates are applied to all cache devices, then data consistency is improved, but input-output operations increase
Solution Approach 1:
Different parts of the cache system receive updates at different times based on their importance. The header and meta-data are updated immediately and mirrored to all cache devices to maintain consistency of control information. Actual data blocks are updated locally without immediate propagation to other devices, allowing asynchronous updates that reduce IO operations while maintaining overall system consistency
3Reliability
If cache blocks are mirrored across devices, then reliability is improved, but storage space efficiency deteriorates
Solution Approach 1:
The cache system separates control data (header and meta-data) from actual data blocks. Only the small control structures are mirrored across multiple cache devices to provide fault tolerance, while the larger data blocks are distributed without mirroring. This segmentation enables reliable operation with minimal redundancy, improving storage space efficiency compared to full mirroring of all data
Solution Approach 2:
An asymmetric redundancy strategy is implemented where critical control information (header and meta-data) is symmetrically mirrored across all cache devices, but actual data blocks are asymmetrically distributed without mirroring. This asymmetric approach provides necessary fault tolerance for system operation while minimizing the storage overhead associated with redundancy
Data Source
AI summary
A data management device includes a cache for a data storage device and a processor. The cache includes cache devices that store a block set. The processor obtains a cache modification request that specifies a first block of the block set, updates a copy of a header of the block set in each of the cache devices based on the modification request, updates a copy of meta-data of the block set in each of the cache devices based on the cache modification request, and updates the first block in a first cache device of the cache devices based on the cache modification request.