Storage System Defragmentation by Reference Frequency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Deduplication storage systems experience a decrease in I/O performance due to fragmentation, which is not efficiently addressed by existing defragmentation techniques that do not consider data content.
Innovation Solution
A storage system that includes a data storage controlling part to track reference frequency of data and a defragmentation processing part to move and store data based on this frequency, optimizing data placement to reduce fragmentation and maintain performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stability of the object's composition
If defragmentation is executed on all regions in a deduplication storage system, then fragmentation is overcome, but I/O performance decreases due to high I/O load
Solution Approach 1:
The patent applies local quality by differentiating defragmentation treatment based on data characteristics. It identifies regions with high duplication rates and exempts them from defragmentation, while only defragmenting regions with low duplication rates. This localized approach optimizes the balance between overcoming fragmentation and maintaining I/O performance by applying defragmentation selectively rather than uniformly across all storage regions.
2Quantity of substance
If deduplication storage system operates for a long period, then storage capacity increases, but free space becomes fragmented and I/O performance decreases
Solution Approach 1:
The patent implements preliminary action by proactively managing free space through selective defragmentation before fragmentation severely impacts performance. It continuously monitors duplication rates and initiates defragmentation operations on low duplication rate regions to maintain free space continuity, preventing the accumulation of fragmentation that would otherwise degrade I/O performance over time.
3Stability of the object's composition
If existing defragmentation techniques are used, then fragmentation is addressed, but data content characteristics are not considered leading to inefficient defragmentation
Solution Approach 1:
The patent applies parameter changes by using duplication rate as a key parameter to determine defragmentation targets. It dynamically identifies regions with low duplication rates and directs defragmentation operations to these specific regions, rather than treating all regions uniformly. This parameter-based approach significantly improves defragmentation efficiency by focusing computational resources on regions that actually benefit from defragmentation.
Data Source
AI summary
A storage system according to the present invention includes: a data storage controlling part that stores data into a storage device and, when storing other data of the same data content as the data, refers to the already stored data as the other data; and a defragmentation processing part that moves and stores storage target data stored in an area set as a defragmentation range within a predetermined region of the storage device, into another region of the storage device. The data storage controlling part stores reference frequency that is frequency of referring to data as other storage target data. The defragmentation processing part, depending on the reference frequency of the data, stores the data into an area to become a defragmentation range later within another region of the storage device.