Mixed Disk RAID Allocation for NVMe and SAS Performance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional RAID storage systems fail to optimize disk allocation, leading to reduced performance due to mixing different types of SSDs, such as NVME and SAS, which affects the overall storage system's capacity and IO performance.
Innovation Solution
A method is proposed to distinguish between different types of disks during RAID set grouping, allocating the same type of disks together to maximize performance, thereby improving storage system efficiency by creating separate RAID resilience sets based on disk types.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If different types of disks (NVME and SAS) are mixed in the same RAID set, then the storage system can maximize disk utilization and increase available capacity, but the IO performance deteriorates due to heterogeneous disk speeds and characteristics
Solution Approach 1:
The patent divides the storage system into separate RAID sets based on disk types. Specifically, it creates a first RAID set for NVME disks and a second RAID set for SAS disks, preventing mixing of different disk types within the same RAID configuration. This segmentation allows each RAID set to optimize for its specific disk type's performance characteristics while still utilizing all available disks to maximize overall capacity.
2Productivity
If traditional RAID allocation methods are used without distinguishing disk types, then the allocation process is simple and fast, but the storage system reliability and performance optimization are compromised
Solution Approach 1:
The patent applies local quality by treating different disk types with different allocation strategies. Instead of uniform treatment, it identifies and separates NVME and SAS disks into distinct RAID sets, allowing each disk type to be allocated according to its specific performance and reliability characteristics. This localized differentiation optimizes both reliability and allocation efficiency without requiring complex global optimization.
Data Source
AI summary
Techniques involve determining the number of disks in a Redundant Array of Independent Disks (RAID) storage system, and the storage system comprises multiple types of mixed disks. The techniques further involve determining a target number of RAID sets to be allocated in the storage system according to the number of disks and a predetermined threshold, and then allocating the multiple types of disks to the RAID sets according to the target number and types of disks. Such techniques propose a way to allocate mixed disks to the corresponding RAID sets. When the RAID set grouping is performed to mixed disks of the storage system, the same type of disks are allocated in the same RAID set as much as possible, thereby improving the performance of the storage system.