Dynamic Storage Hierarchy Management for Data Access Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional storage systems are limited by relying on a single type of storage medium, which can lead to inefficiencies due to the inherent advantages and disadvantages of that medium, such as high costs for high-performance flash memory and slow performance of hard disk drives, without effectively leveraging the strengths of multiple memory types.
Innovation Solution
A dynamic storage allocation architecture that monitors data access patterns and attributes to optimize storage by distributing data across multiple memory types, such as flash memory and hard disk drives, based on performance characteristics like I/O operations, cost, and volatility, allowing for real-time re-allocation to match changing usage demands.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If flash memory is used for storage, then response time is improved and I/O operations per second are increased, but cost per bit stored increases
Solution Approach 1:
The patent segments the storage system into multiple storage devices with different performance characteristics (e.g., flash memory for high-speed operations, hard disk drives for capacity). Data is divided and placed in appropriate segments based on access patterns, allowing the system to achieve high response times for frequently accessed data while using lower-cost storage for less critical data.
Solution Approach 2:
The patent applies local quality by assigning different storage media to different data based on their specific requirements. High-performance flash memory is allocated to data requiring fast access, while lower-cost hard disk drives are used for data with less stringent performance requirements. This creates locally optimized storage quality matching data needs.
2Ease of manufacture
If hard disk drive is used for storage, then cost per bit stored is reduced, but response time deteriorates and sensitivity to harsh operation conditions increases
Solution Approach 1:
The system segments storage functions by using hard disk drives for data that does not require rapid access, thereby reducing overall system cost while maintaining acceptable performance for non-critical data. This segmentation allows the system to leverage the cost advantages of HDDs without compromising the performance of the entire storage system.
Solution Approach 2:
The patent introduces a storage management component that acts as an intermediary, intelligently routing data between hard disk drives and flash memory based on access patterns. This mediator ensures that data requiring fast access is automatically placed in flash memory, while other data remains on the cost-effective hard disk drive.
3Device complexity
If data is stored in a single medium, then storage architecture is simplified, but system performance and cost-effectiveness are limited by the medium's inherent characteristics
Solution Approach 1:
The patent creates a universal storage system that can handle multiple types of data with different performance requirements using a single multi-functional architecture. The system can dynamically allocate data across different storage media types, making the storage system adaptable to various data needs without requiring separate dedicated storage systems for each data type.
Solution Approach 2:
The patent implements dynamic data placement and migration capabilities, allowing the storage system to adaptively move data between different media types based on changing access patterns. This dynamic behavior enables the system to optimize performance in real-time while maintaining a relatively simple underlying storage architecture.
Data Source
AI summary
A data block may be moved between a first medium and a second medium. The movement of the data block involves measuring the access characteristic of the data block as the data block is stored on the first medium. The performance characteristics of the first medium and the second medium are then determined, in which each performance characteristic has a static performance characteristic component and a dynamic performance characteristic component. Alternatively or concurrently, the static performance characteristic components of the first medium and the second medium may be compared, and the dynamic performance characteristic components of the first medium and the second medium are compared. Accordingly, the data block is moved from the first medium to the second medium when at least one of these comparisons indicate that the second medium is more suitable for storing the data block having the access characteristic than the first medium.