Dynamic Storage Capacity Allocation in Software-Defined Storage

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing storage provision services using software-defined storage (SDS) face inefficiencies in capacity utilization, as some tenants use more physical storage while others use less, leading to reduced usage efficiency due to unadjusted overcommit ratios.

Innovation Solution

An operation management system that adjusts physical capacity allocation within logical volumes based on data size written by tenants, using SDS nodes and a capacity allocation and recovery system to dynamically redistribute physical capacities among tenants, maintaining optimal overcommit ratios.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a fixed overcommit ratio is allocated to each tenant, then service level agreements are maintained, but storage usage efficiency is reduced

Engineering Contradiction:
Improveservice level agreementVSAvoidstorage usage efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent implements dynamic adjustment of physical capacity allocation based on actual data write amounts. The system continuously monitors the ratio between logical volume size and physical capacity allocation, and automatically adjusts the physical capacity when this ratio deviates from the predetermined overcommit ratio range, transforming the static allocation into a dynamic adaptive system that maintains both service level agreements and storage efficiency

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system establishes a feedback mechanism where the physical capacity allocation is continuously monitored and adjusted based on the actual data write amount. When the ratio of logical volume size to physical capacity exceeds or falls below the overcommit ratio range, the system automatically adjusts the physical capacity allocation to bring the ratio back within the acceptable range, ensuring both reliability and efficiency

Inventive Principle:
Principle #23Feedback

2Ease of operation

If physical capacity is allocated based on logical volume size with fixed overcommit ratio, then capacity planning is simplified, but actual storage utilization becomes inefficient

Engineering Contradiction:
Improvecapacity planningVSAvoidstorage utilization
Core Design Contradiction:
Ease of operationVSProductivity

Solution Approach 1:

The system implements self-service automatic adjustment where the storage system autonomously monitors its own capacity allocation and dynamically adjusts physical capacity based on actual data write amounts. This eliminates the need for manual capacity planning while maintaining efficient utilization, as the system automatically recalibrates the physical capacity to match actual usage patterns

Inventive Principle:
Principle #25Self-service

Data Source

PatentUS10705876B2Operation management system and operation management method
Publication Date: 2020.07.07 HITACHI VANTARA LTD
  • US10705876B2 patent drawing
  • US10705876B2 patent drawing
  • US10705876B2 patent drawing

AI summary

An SDS node adjusts a physical capacity included in a logical volume provided to a tenant and is to be allocated to the tenant, based on the size of data written from the tenant. An operation management system provides storage to the tenant and adjusts the provided storage to include a node and a capacity allocation and recovery system. The node provides the logical volume of the storage to the tenant, receives writing of data to the logical volume from the tenant, and writes the received data to a region included in the storage and having the physical capacity allocated to the tenant. The capacity allocation and recovery system determines a physical capacity to be interchanged based on the size of the logical volume, the size of the written data, and the physical capacity allocated to the tenant, and notifies information on the determined physical capacity to the node.