Dynamic Storage Tiering via Block-Level Cache Hit Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data tiering systems face inefficiencies in managing data access patterns, particularly in file-based solutions that fail to relocate un-accessed portions of files to lower performing storage tiers, leading to suboptimal performance and capacity utilization.
Innovation Solution
A block-based dynamic storage tiering method that tracks cache hits for logical block addresses to determine data access activity, allowing for the relocation of data between high and low performance storage tiers, thereby optimizing storage allocation based on actual usage patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If file-based tiering solutions are used, then data movement can be achieved, but un-accessed portions of files cannot be relocated to lower performing storage tiers
Solution Approach 1:
The patent segments files into smaller data units (e.g., data blocks, extents, or stripes) that can be independently managed and relocated. This allows only the frequently accessed portions of files to remain in high-performance storage while un-accessed portions are moved to lower-performance storage tiers, resolving the limitation of file-based tiering that treats entire files as atomic units.
2Speed
If data is kept in high-performance storage, then access speed is improved, but storage capacity is wasted on infrequently accessed data
Solution Approach 1:
The patent implements dynamic data placement that continuously monitors access patterns and automatically relocates data between storage tiers based on actual usage. Data that was previously statically placed in high-performance storage is now dynamically moved to lower-performance storage when access patterns indicate it is infrequently accessed, optimizing both speed and capacity utilization.
3Adaptability or versatility
If host-based software is used for data movement, then application awareness is achieved, but system complexity increases
Solution Approach 1:
The patent enables the storage system to autonomously monitor its own access patterns and perform self-optimization through automatic data relocation. The storage array's controller monitors I/O operations and cache hit rates, then automatically relocates data without requiring external host-based software, reducing overall system complexity while maintaining adaptability.
Data Source
AI summary
A method for dynamic storage tiering may include, but is not limited to: receiving an input/output (I/O) request from a host device; determining whether the I/O request results in a cache hit; and relocating data associated with the I/O request between a higher-performance storage device and lower-performance storage device according to the determination whether the data associated with the I/O request is stored in a cache.