Tiered Storage Controller Dynamic Data Placement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current tiered data storage systems in information handling systems lack efficient mechanisms to dynamically manage data placement based on access frequency, leading to suboptimal performance and resource utilization.
Innovation Solution
A tiered data storage system with a storage controller that uses metadata structures to track access frequencies and logically divide data into chunks, allowing for dynamic movement of data between storage tiers to optimize access times, thereby enhancing application performance by focusing scans and archiving operations on frequently and infrequently accessed data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If data is placed in fast storage media, then access speed is improved, but storage cost increases
Solution Approach 1:
The storage system is divided into multiple tiers with different performance and cost characteristics. Frequently accessed data is segmented and placed in fast storage media, while less frequently accessed data is segmented and placed in slower, cheaper storage media. This segmentation allows the system to optimize both access speed for critical data and cost for bulk storage.
Solution Approach 2:
Different portions of the storage system are assigned different qualities based on local access patterns. Hot data regions receive high-performance storage resources, while cold data regions use cost-effective storage resources. This local quality differentiation ensures that fast storage is applied only where needed, reducing overall cost while maintaining necessary access speeds.
2Productivity
If all data is scanned for access frequency tracking, then data placement optimization is improved, but scan time increases
Solution Approach 1:
Instead of scanning all data uniformly, the system performs partial scanning focused on specific regions or data types that are more likely to exhibit access pattern changes. Metadata structures track access frequencies for only the most relevant data portions, providing sufficient optimization information without the overhead of comprehensive scanning of entire datasets.
Solution Approach 2:
The system preliminarily identifies and tracks access patterns for data that is likely to benefit from tiered placement based on metadata characteristics or access history. By pre-identifying candidate data for frequency tracking before full scanning occurs, the system reduces unnecessary scan operations and focuses computational resources on data that will provide the most benefit from optimization.
Data Source
AI summary
An information handling system includes a memory controller coupled to a first and a second storage device. The memory controller stores information on the first storage device based upon an access count related to the information, stores other information on the second storage device based upon a another access count related to the other information, and stores access data including the access counts. The information handling system also includes a processing system coupled to the memory controller. The processing system obtains the access data from the memory controller, and determines to perform an operation on the data based upon the access count, and to not perform the operation on the other data based upon the other access count.