Disk Slice Tiering via Multi-Dimensional Performance Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional Virtual Pool Fully Automated Storage Tiering (FAST VP) technology only considers access frequency and not performance information of disk slices, leading to inefficient data movement and suboptimal cost-performance ratios in tiered storage operations.
Innovation Solution
A method to determine performance information of disk slices based on differences in performance parameters such as bandwidth, IOPS, and response time, and calculate comprehensive scores combining access frequency and performance information to accurately evaluate and reposition disk slices across different storage tiers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional FAST VP technology only considers access frequency for data tiering, then the implementation is simple and fast, but the data movement efficiency is low and cost-performance ratio is suboptimal
Solution Approach 1:
The patent changes the evaluation parameters from only access frequency to a multi-dimensional parameter set including access frequency, read IOPS, write IOPS, and response time. This allows the system to make more informed tiering decisions by considering performance characteristics alongside access patterns, thereby improving data movement efficiency while maintaining manageable complexity through structured parameter collection and scoring.
2Reliability
If performance information of disk slices is not considered, then the tiering process is simple and fast, but the cost-performance ratio is suboptimal
Solution Approach 1:
The patent performs preliminary action by pre-collecting and caching performance information (read IOPS, write IOPS, response time) for different disk types before actual tiering decisions are made. This preparation work is done in advance so that when tiering decisions need to be made, the system can quickly retrieve and compare pre-computed performance data alongside access frequency, reducing the time penalty of considering multiple parameters while improving tiering optimization reliability.
3Productivity
If only access frequency is used for disk slice evaluation, then the evaluation process is simple and quick, but the data placement optimization is insufficient
Solution Approach 1:
The patent segments the evaluation process into distinct components: access frequency evaluation, performance information evaluation (read IOPS, write IOPS, response time), and a综合 scoring system that combines these segments with weighted coefficients. This segmentation allows each aspect to be evaluated independently and then integrated, making the complex scoring system more manageable and enabling better storage efficiency through comprehensive evaluation without overwhelming complexity.
Data Source
AI summary
Techniques for managing disks involve determining performance information of an access pattern of a disk slice based on differences in performance parameters of the access pattern of the disk slice on a plurality of disks. Such techniques further involve determining a score for the disk slice based on the performance information and access frequency information of the disk slice. Such techniques further involve determining a position of the disk slice in the plurality of disks based on the score.