Elastic Cache Extents for Redundant Data Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional storage systems inefficiently use Flash module space due to redundant caching methods, which reduce available cache space and do not effectively utilize existing extent structures and management infrastructures.
Innovation Solution
An elastic caching method that dynamically allocates and deallocates cache extents based on write requests, using existing extent structures and management infrastructures, and implements various RAID levels for redundancy, reducing buffer writes and enhancing Flash memory life by selectively mirroring dirty writes across multiple nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If mirrored cache lines are associated via pointers in a dedicated data structure, then data redundancy is provided, but cache space utilization deteriorates
Solution Approach 1:
The patent segments the cache into multiple extents that can be independently managed and dynamically allocated. Instead of dedicating entire cache modules to mirroring, the cache is divided into smaller units (extents) that can be flexibly assigned to different data blocks, allowing partial mirroring and better space utilization while maintaining redundancy where needed.
Solution Approach 2:
The patent implements dynamic mirroring where the mirroring ratio and extent associations are not fixed but can be adjusted based on workload characteristics, data access patterns, and available cache space. This allows the system to optimize between redundancy and space utilization dynamically, rather than being constrained by static pointer-based associations.
2Reliability
If two Flash modules are permanently paired using mirroring, then data protection is improved, but available cache space is reduced
Solution Approach 1:
The patent applies mirroring selectively to specific extents and data blocks rather than uniformly across all cache modules. Different regions of the cache can have different mirroring ratios based on the importance and access patterns of the stored data, allowing critical data to be protected while less critical data uses fewer resources.
Solution Approach 2:
The patent enables cache extents to serve multiple functions - they can be used for primary cache storage, mirrored for redundancy, or allocated to different data blocks as needed. The extent structure is designed to be versatile, supporting various RAID levels and mirroring configurations within the same cache infrastructure.
3Ease of operation
If cache-wide basis selection for mirroring is made, then implementation simplicity is improved, but fine-grained space optimization is lost
Solution Approach 1:
The patent segments the cache into extents that can be independently managed, allowing fine-grained control over which portions of the cache are mirrored. This segmentation enables selective mirroring at the extent level rather than requiring cache-wide decisions, optimizing space usage while maintaining manageable complexity through the use of extent pointers and metadata structures.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
apparatus for elastic caching of redundant cache data. The apparatus may have a plurality of buffers and a circuit. The circuit may be configured to (i) receive a write request from a host to store write data in a storage volume, (ii) allocate a number of extents in the buffers based upon a redundant organization associated with the write request and (iii) store the write data in the number of extents, where (a) each of the number of extents is located in a different one of the buffers and (b) the number of extents dynamically linked together in response to the write request.