Dynamic Data Allocation via Performance-to-Capacity Ratios
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In distributed storage systems, the existing method of allocating data based solely on disk capacity leads to suboptimal performance, as disks with higher capacities often have the same input/output operations per second (IOPS) as those with lower capacities, resulting in underutilization of storage performance and capacity.
Innovation Solution
A method and apparatus that dynamically adjust data allocation shares among disks by determining a reference value based on their performance-to-capacity ratios, allowing for reallocation of data to optimize storage performance by prioritizing disks with higher performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data is allocated according to capacity proportions of disks, then storage capacity utilization is improved, but storage performance is severely affected because smaller-capacity disks fill up first while larger-capacity disks remain underutilized
Solution Approach 1:
The patent changes the allocation parameter from capacity proportion to performance-to-capacity ratio. By calculating this ratio for each disk and using it as the basis for data allocation, the system ensures that disks with better performance relative to their capacity receive more data, thereby balancing both capacity utilization and performance optimization
Solution Approach 2:
The patent implements dynamic allocation by continuously monitoring storage status and recalculating allocation shares based on current performance-to-capacity ratios. When a disk reaches full-write state, the system dynamically adjusts allocation by emigrating data from full disks to disks with available space, ensuring optimal performance throughout the storage lifecycle
2Quantity of substance
If more data is allocated to larger-capacity disks, then storage capacity is fully used, but storage performance cannot be fully brought into play because smaller-capacity disks with same IOPS fill up first
Solution Approach 1:
The patent introduces a new parameter - performance-to-capacity ratio - that combines both capacity and performance metrics. This parameter transformation allows the system to allocate data in a way that simultaneously considers both the size and performance characteristics of each disk, preventing the performance utilization problem
Solution Approach 2:
The patent creates equipotential allocation by ensuring that all disks contribute equally to system performance through the performance-to-capacity ratio mechanism. Each disk's allocation share is proportional to its performance contribution relative to its capacity, creating a balanced system where no disk is over or under-utilized
Data Source
AI summary
In a method for allocating data to multiple disks for storage, a capability value of each of the multiple disks is obtained. The capability value is ratio of a performance indicator value of a disk to a maximum capacity value of the disk. A greatest capability value is identified from multiple capability value. Based on the greatest capability value and a maximum capacity value of each of the multiple disks, allocation shares of the multiple disks are determined. Based on the determined allocation shares, data is allocated for the multiple disks. Through the method, the overall storage performance of a storage system is improved.