Managing Apparatus for Predictive Storage Load Balancing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional systems experience high Input/Output (I/O) loads on shared disks due to excessive access frequencies from multiple apparatuses, leading to increased load before and during data copying processes, which is inefficient in managing storage access loads.

Innovation Solution

A managing apparatus with a predicting unit that identifies access patterns and periodicity in storage access frequencies, and a controlling unit that proactively transfers data from high-access areas to other storage areas before the access load becomes excessive, thereby reducing I/O loads on primary storage devices.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If the entirety of the storage contents of the first disk is copied onto the second disk after the access frequency exceeds a threshold value, then the access load on the first disk is reduced, but the I/O load on the first disk becomes even larger during the copying process

Engineering Contradiction:
Improveaccess load reductionVSAvoidI/O load during copying
Core Design Contradiction:
ProductivityVSLoss of energy

Solution Approach 1:

The predicting unit forecasts future access frequency trends based on historical access patterns, and the controlling unit proactively copies data to the second disk before the access frequency actually exceeds the threshold. This preliminary action prevents the high I/O load problem by performing copying during low-access periods rather than after access frequency becomes excessive.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system continuously monitors access frequency to the first disk and uses this feedback information to trigger copying operations. The predicting unit analyzes access patterns and provides predictive feedback, enabling the controlling unit to make informed decisions about when to copy data, thereby optimizing the timing to avoid high I/O loads.

Inventive Principle:
Principle #23Feedback

2Ease of operation

If control is exercised only after access frequency exceeds a threshold value, then the system responds to actual load conditions, but the I/O load on the first disk is already large before control is exercised

Engineering Contradiction:
Improvereactive control simplicityVSAvoidtime with high I/O load
Core Design Contradiction:
Ease of operationVSLoss of time

Solution Approach 1:

Instead of waiting for access frequency to exceed the threshold, the predicting unit analyzes historical access patterns and forecasts future access frequency. When the prediction indicates that access frequency will exceed the threshold, the controlling unit proactively initiates copying operations during low-access periods, thereby eliminating the time period with high I/O load.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses its own access frequency data and patterns to predict future load conditions and automatically trigger copying operations. This self-service approach enables the system to optimize its own performance by performing maintenance operations at optimal times without external intervention.

Inventive Principle:
Principle #25Self-service

3Adaptability or versatility

If multiple apparatuses access a shared volume storing OS programs and application programs, then resource sharing is achieved, but the I/O load on the shared volume becomes large

Engineering Contradiction:
Improveresource sharingVSAvoidI/O load on shared volume
Core Design Contradiction:
Adaptability or versatilityVSLoss of energy

Solution Approach 1:

The system segments the storage load by dividing data between the first disk (shared volume) and the second disk (local copy). The controlling unit determines which data to copy to the second disk based on access frequency predictions, thereby segmenting the I/O load and reducing the burden on the shared volume while maintaining resource sharing capabilities.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The second disk acts as an intermediary between the multiple apparatuses and the shared volume. By copying frequently accessed data to the second disk, the system reduces direct access to the shared volume, thereby reducing I/O load while still enabling multiple apparatuses to access the data through the intermediary.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentUS9501428B2Managing apparatus
Publication Date: 2016.11.22 FUJITSU LTD
  • US9501428B2 patent drawing
  • US9501428B2 patent drawing
  • US9501428B2 patent drawing

AI summary

A managing server includes a predicting unit and a controlling unit. The predicting unit predicts either a time or a time period at which frequency of accesses from a plurality of apparatuses to a storage device storing data therein will exceed a predetermined threshold value, on a basis of a status of the accesses from the plurality of apparatuses to the storage device. The controlling unit controls the storage device so as to transmit the data stored therein to at least one of the plurality of apparatuses, prior to the predicted time or time period.