Deferred Copy Target Pull for Distributed Storage Efficiency

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In composite storage servers, the significant processing power required by single source copy sites and copy target sites to manage and distribute volume data efficiently is not effectively balanced, leading to inefficiencies due to the exclusion of copy target sites from decision-making processes and concurrent execution of storage jobs.

Innovation Solution

Implementing a deferred copy target site that pulls volume data from a peer storage site only after receiving a source copy complete message and detecting the host job's transition to an inactive state, allowing for active participation of all distributed storage sites in data storage.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If a single source copy site pulls and pushes cached volume data to multiple copy target sites, then the volume data can be distributed to all sites, but the source copy site demands significant processing power and creates bottlenecks

Engineering Contradiction:

Improvevolume data distribution efficiencyVSAvoidprocessing power demand at source copy site

Core Design Contradiction:

ProductivityVSPower

Solution Approach 1:

The patent segments the centralized data distribution task into multiple distributed pull operations. Instead of one source site pushing data to all targets, each copy target site independently pulls data from peer sites that have complete copies. This distributes the processing load across multiple sites rather than concentrating it at a single source.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Copy target sites are empowered to make their own decisions about when, where, and how fast to pull volume data. Each site autonomously monitors the active/inactive state of host jobs and initiates data pulls from appropriate peer sites without requiring centralized coordination, thereby serving itself and reducing the burden on any single source site.

Inventive Principle:

Principle #25Self-service

2Ease of operation

If copy target sites are excluded from decision-making processes, then the source copy site has full control over data distribution, but copy target sites also demand significant processing power to handle concurrent storage jobs

Engineering Contradiction:

Improvecentralized control simplicityVSAvoidprocessing power demand at copy target sites

Core Design Contradiction:

Ease of operationVSPower

Solution Approach 1:

The patent enables copy target sites to autonomously determine when to pull data by monitoring the active/inactive state of host jobs. Each site independently makes decisions about data retrieval timing and source selection, eliminating the need for complex centralized coordination while reducing processing demands through distributed intelligence.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

Instead of the source copy site pushing data to targets, the patent inverts the approach by having copy target sites actively pull data from peers. This reverses the traditional client-server dynamic, allowing targets to control their own data acquisition and reducing the operational burden on source sites.

Inventive Principle:

Principle #13The other way round (Inversion)

3Productivity

If all distributed storage sites actively participate in data storage decisions, then processing burdens are distributed and efficiency is enhanced, but the system complexity increases

Engineering Contradiction:

Improvestorage operation efficiencyVSAvoiddistributed decision-making complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent implements a uniform rule across all distributed storage sites: each site monitors the active/inactive state of host jobs and pulls data from peers when appropriate. This homogeneous approach simplifies the system architecture compared to complex centralized control, as all sites follow the same simple logic rather than requiring sophisticated coordination protocols.

Inventive Principle:

Principle #33Homogeneity

Data Source

PatentUS8019723B2Deferred copy target pull of volume data

Publication Date: 2011.09.13 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

A composite storage server employs a plurality of distributed storage sites including a mount site, a source copy site and one ore more copy target sites for executing a storage of volume data within the composite storage server. In operation, a deferred copy target site receives a communication of a source copy complete message from the mount site, the source copy complete message being indicative of the source copy site having a complete copy of the volume data, and detects a transition by the mount site of the host job from an active status to an inactive status subsequent to the communication of the source copy complete message from the mount site. Based on the inactive status of the host job, the deferred copy target site pulls the volume data from a peer distributed storage site having a complete copy of the volume data.