Storage Allocation Uniformity for RAID Rebuilding Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional RAID technologies face challenges in reducing rebuilding time due to increased risk of double disk failure with new technologies like shingled media disks, where the write input/output bandwidth of the hot spare disk becomes a bottleneck, leading to potential data loss.
Innovation Solution
A method and device for storage management that determines and selects allocation schemes to minimize allocation uniformity variation, ensuring even distribution of storage devices and reducing rebuilding time by using a subset of candidate schemes that cover all storage devices, optimizing allocation status and size considerations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If disk capacity is increased using new technologies like shingled media disks, then storage capacity is improved, but rebuilding time increases and risk of double disk failure increases
Solution Approach 1:
The patent segments the hot spare disk into multiple virtual disks, allowing parallel rebuilding operations. When a disk fails, the system can simultaneously rebuild data across multiple virtual disk targets, significantly reducing the total rebuilding time compared to sequential rebuilding on a single hot spare disk.
Solution Approach 2:
The patent introduces a virtualization dimension to the physical hot spare disk by creating multiple virtual disk instances. This transforms the single-disk sequential rebuilding process into a multi-disk parallel rebuilding process, effectively adding a dimensional layer to overcome the capacity-time tradeoff.
2Quantity of substance
If disk capacity is increased using new technologies like shingled media disks, then storage capacity is improved, but reliability deteriorates due to increased risk of double disk failure
Solution Approach 1:
By segmenting the hot spare into multiple virtual disks, the system can distribute rebuilding operations across multiple targets. This segmentation allows the system to maintain multiple independent rebuild paths, reducing the probability that a single point of failure will cause complete rebuild failure.
Solution Approach 2:
The patent prepares multiple virtual hot spare disks in advance, creating redundant rebuild paths before failures occur. This beforehand preparation ensures that if one rebuild path fails due to another disk failure during the process, alternative paths are already available to complete the rebuild.
3Reliability
If conventional RAID technology is used, then data redundancy is maintained, but write input/output bandwidth becomes a bottleneck for reducing rebuilding time
Solution Approach 1:
The patent merges multiple virtual hot spare disks into a unified virtualization layer that manages parallel rebuilding operations. This merging allows the system to combine the write bandwidth of multiple physical disks, effectively multiplying the rebuilding throughput while maintaining RAID data redundancy through the virtualization management layer.
Solution Approach 2:
The virtualization layer introduces a new operational dimension that enables parallel write operations across multiple virtual disk targets simultaneously. This dimensional transformation allows the system to overcome the single-disk write bandwidth bottleneck by distributing write operations across multiple physical disks in parallel.
Data Source
AI summary
Embodiments of the present disclosure provide a method and device for storage management. For example, there is proposed a method comprising: in response to a plurality of storage devices in a storage system being to be allocated to an unallocated logic storage area, determining a plurality of allocation schemes for allocating the plurality of storage devices to the unallocated logic storage area; obtaining allocation uniformity of the plurality of storage devices with respect to an allocated logic storage area of the storage system; and selecting one of the plurality of allocation schemes at least based on the allocation uniformity, such that the uniform degree of the allocation has a minimum variation. Corresponding device and computer program product are also disclosed.