Virtual Storage Unit Mapping for I/O Load Balancing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current storage systems face performance issues due to I/O load imbalances among disks, as physical storage units are often associated with only a few disks, leading to uneven access frequencies and reduced overall system performance.
Innovation Solution
The solution involves mapping physical storage units evenly across multiple disks into virtual storage units, distributing data in a round-robin manner across these units, and managing storage space allocation, access requests, and data migration in units of virtual storage units, while handling storage expansion and disk failures in units of physical storage units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If physical storage units are allocated on demand and associated with only a few disks, then storage space management is simplified, but I/O load becomes unbalanced among disks
Solution Approach 1:
The patent segments the storage system into multiple levels: physical storage units at the disk level, virtual storage units as the basic management unit, and storage pools as the highest level. This segmentation allows virtual storage units to be flexibly allocated across multiple physical storage units, enabling both simplified management and balanced I/O distribution. The virtual storage unit acts as an intermediate layer that decouples the simplicity of management from the complexity of load balancing.
Solution Approach 2:
The patent introduces virtual storage units as an intermediary between the storage management system and physical disks. This intermediary layer enables the system to manage storage space uniformly through virtual units while simultaneously distributing I/O operations across multiple physical disks in a balanced manner. The virtual storage unit mapping mechanism serves as the mediator that resolves the contradiction between management simplicity and load balance.
2Speed
If data in certain physical storage units is accessed frequently, then quick access to that data is achieved, but I/O load imbalance occurs among disks
Solution Approach 1:
The patent applies local quality by allowing different virtual storage units to be mapped to different physical storage units based on access patterns and performance requirements. Hot data can be placed on faster disks while cold data resides on slower disks, optimizing local access speed without compromising overall system performance. This selective mapping ensures that frequently accessed data benefits from high-speed access while maintaining balanced I/O distribution across the entire system.
Solution Approach 2:
The patent implements dynamic mapping between virtual and physical storage units, allowing the system to adapt to changing access patterns. When access patterns change, the mapping can be reconfigured to maintain optimal performance. This dynamic adjustment enables the system to respond to varying data access requirements while preserving overall I/O balance and system performance.
Data Source
AI summary
Techniques involve: in response to receiving an access request for a logical address, determining a first virtual storage unit and a first offset within the first virtual storage unit corresponding to the logical address; determining a first set of virtual storage units to which the first virtual storage unit belongs and a position of the first virtual storage unit, determining, from the plurality of physical storage units corresponding to the first set of virtual storage units and based on the position and the first offset, a physical storage unit and a second offset within the physical storage unit corresponding to the logical address; and performing, based on an identifier of the physical storage unit and the second offset, a data access operation requested by the access request. Accordingly, I/O loads on a plurality of disks can be balanced, thereby improving overall performance of a storage system.