Storage Resource Headroom Prediction via Latency Utilization Curves

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In networked storage environments, managing resources efficiently to maintain optimal performance capacity is challenging due to varying workloads and complexities in determining available headroom, leading to potential overestimation of performance capacity and misallocation of resources.

Innovation Solution

The implementation of a headroom module that collects QOS data to determine available performance capacity by analyzing the relationship between latency and utilization, categorizing workloads based on service time and variability, and using a model-based technique to generate the latency vs. utilization curve, allowing for real-time prediction of optimal resource utilization.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If resource performance capacity is estimated without accurate measurement, then resource allocation can be simplified, but performance capacity is overestimated leading to inefficient resource management

Engineering Contradiction:

Improveresource allocation simplicityVSAvoidperformance capacity measurement accuracy

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The patent replaces complex manual resource capacity estimation with an automated electronic system that collects QOS data, generates latency-utilization curves, and calculates headroom programmatically. This substitution of manual assessment with automated computational methods resolves the contradiction by providing precise measurements without increasing operational complexity for users.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system performs self-measurement of resource performance capacity by automatically collecting its own QOS data, generating latency vs. utilization curves, and calculating available headroom without external intervention. This self-service capability enables accurate measurement while maintaining ease of operation, as the system autonomously resolves its own performance assessment needs.

Inventive Principle:

Principle #25Self-service

2Productivity

If resource utilization is increased to maximize throughput, then system capacity is improved, but performance capacity may be overestimated and service level objectives may not be met

Engineering Contradiction:

Improvesystem throughput capacityVSAvoidservice level objective compliance

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system continuously monitors QOS data including latency and utilization metrics, generates updated latency vs. utilization curves, and recalculates headroom based on current system state. This closed-loop feedback mechanism enables dynamic adjustment of resource allocation to maintain both high throughput and reliable service level compliance by preventing overestimation of available capacity.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent implements dynamic resource capacity assessment through continuously updated latency vs. utilization curves that reflect changing system conditions. Rather than static capacity estimates, the system adapts headroom calculations to current workload patterns and performance characteristics, enabling both high productivity and reliable SLO compliance under varying operational conditions.

Inventive Principle:

Principle #15Dynamics

3Measurement precision

If latency vs. utilization curve analysis is implemented to accurately determine headroom, then performance capacity measurement precision is improved, but system complexity increases

Engineering Contradiction:

Improveavailable performance capacity measurement accuracyVSAvoidsystem architectural complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The headroom module serves multiple functions: collecting QOS data, generating latency vs. utilization curves, calculating headroom, and providing recommendations for resource allocation. By consolidating these diverse functions into a single multi-functional module, the system achieves precise performance measurement without proportionally increasing overall system complexity.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent introduces a headroom module as an intermediary component that bridges raw QOS data collection and resource allocation decisions. This intermediary layer processes complex latency vs. utilization curve analysis and translates it into actionable headroom metrics, shielding the rest of the system from complexity while maintaining measurement precision.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS10048896B2Methods and systems for determining performance capacity of a resource of a networked storage environment

Publication Date: 2018.08.14 NETAPP INC

AI summary

Methods and systems for a networked storage system are provided. One method includes categorizing by a processor performance data associated with a resource used in a networked storage environment for reading and writing data at a storage device based on a workload mix, where the workload mix is determined by a service time in which the resource processes the workload mix, a parameter indicating variability of the service time and a utilization bin index value indicating resource utilization at a given time; and determining by the processor available performance capacity of the resource using the categorized performance data, where the available performance capacity is based on optimum utilization of the resource and utilization of the resource.