I/O Classification-Based Data Movement in Storage Pools

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current data storage systems face challenges in optimizing data movement across storage pools based on performance goals and workload characteristics, leading to inefficiencies in response time and resource utilization.

Innovation Solution

A method is introduced to determine data movements by classifying workloads into specific I/O classifications and matching them with preferred storage pools based on performance goals and capacity limits, using a matrix to model expected performance and adjust data placement between storage tiers to optimize response time objectives.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of time

If data is stored on high-performance storage pools to meet response time objectives, then response time is improved, but storage cost increases

Engineering Contradiction:

Improveresponse timeVSAvoidstorage cost

Core Design Contradiction:

Loss of timeVSQuantity of substance

Solution Approach 1:

The patent applies local quality by assigning different storage pools to different data portions based on their specific I/O characteristics. Hot data with random I/O patterns is placed on high-performance storage pools, while cold data with sequential patterns is placed on lower-performance pools, optimizing response time for critical operations without unnecessarily increasing storage costs across the board.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The system dynamically adjusts data placement between storage pools based on changing workload characteristics. As I/O patterns evolve, the system reclassifies data portions and relocates them to appropriate storage pools, ensuring response time objectives are met while adapting to varying storage cost conditions.

Inventive Principle:

Principle #15Dynamics

2Productivity

If data movement decisions are made without considering I/O classifications, then device complexity is reduced, but productivity decreases

Engineering Contradiction:

Improvedata storage efficiencyVSAvoiddata movement determination complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent changes the parameter of data characterization by introducing I/O classifications that capture essential workload characteristics such as read/write ratios, sequential/random patterns, and operation sizes. This parameter transformation enables more effective data movement decisions without requiring overly complex analysis, improving storage efficiency while maintaining manageable system complexity.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS8949483B1Techniques using I/O classifications in connection with determining data movements

Publication Date: 2015.02.03 EMC IP HLDG CO LLC

AI summary

Techniques are described for determining data movements. A first plurality of performance goals for a plurality of storage pools and a second plurality of performance goals for a plurality of applications are received. A first I/O classification characterizing a workload of a first data portion is determined. The first I/O classification is one of a predetermined set of I/O classifications. A proposed data movement of the first data portion from a first to a second of the plurality of storage pools is determined in accordance with criteria including a match between the first I/O classification and one of the predetermined set of I/O classifications which is preferred for the second storage pool over one or more other I/O classifications of the predetermined set. The criteria also includes any of the first plurality of performance goals and the second plurality of performance goals.