Virtual Storage Load Balancing via IO Limitation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing data storage systems face inefficiencies in load balancing due to limited input-output (IO) capacity of physical storage devices, leading to suboptimal data storage rates when multiple virtual storage devices share the same physical resources.

Innovation Solution

The implementation of virtual storage devices hosted by physical storage devices, which group IO-bound storage devices to distribute file blocks across different physical storage devices, thereby load balancing data storage and improving storage rates by avoiding simultaneous usage limitations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If multiple virtual storage devices share the same physical storage device, then resource utilization is improved, but data storage rate deteriorates due to IO capacity limitations

Engineering Contradiction:

Improveresource utilizationVSAvoiddata storage rate

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The system segments virtual storage devices into different groups based on their physical storage device assignments. When distributing file blocks, the system selects virtual storage devices from different physical storage device groups to avoid IO bottlenecks, thereby maintaining both high resource utilization and high data storage rates

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system introduces a new dimension of distribution by considering physical storage device identity as an additional constraint layer. Instead of only distributing blocks across virtual storage devices, it distributes across groups of virtual storage devices that map to different physical storage devices, effectively adding a spatial dimension to the load balancing strategy

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Productivity

If file blocks are distributed across multiple virtual storage devices, then load balancing is improved, but queueing issues occur in rate-limited virtual storage devices

Engineering Contradiction:

Improveload balancingVSAvoidqueueing time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The system implements feedback mechanisms to monitor the IO status and queue depth of virtual storage devices. When distributing file blocks, it uses this feedback information to dynamically select target virtual storage devices, avoiding those with high queue depths and thereby reducing queueing time while maintaining load balancing

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system performs preliminary assessment of virtual storage device availability and IO capacity before distributing file blocks. By pre-evaluating which virtual storage devices can accept blocks without causing queueing delays, it proactively avoids time loss while achieving load balancing

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS10459633B1Method for efficient load balancing in virtual storage systems

Publication Date: 2019.10.29 EMC IP HLDG CO LLC

AI summary

A data storage device includes virtual storage devices, hosted by physical storage devices, and a processor. The processor obtains a data storage request, divides a file specified by the data storage request into blocks, and stores the blocks in the virtual storage devices based on an input output (IO) limitation of the virtual storage devices.