Storage Traffic Modeling via Neural Network Trend Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current storage traffic modeling in Software Defined Storage (SDS) systems lacks the ability to accurately predict future workload demands, leading to inefficiencies and potential service disruptions due to fixed infrastructure limitations, especially in cloud services where increasing client demands outpace existing hardware capabilities.
Innovation Solution
A method and system utilizing an artificial neural network for storage traffic modeling, which collects observed performance parameters like IOPS, latency, and throughput, learns trends over time, and adjusts predictions to improve accuracy by comparing predicted and observed values, allowing for flexible infrastructure adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If fixed infrastructure is prepared for maximum expected capacity, then service quality can be maintained under peak demand, but resource utilization efficiency deteriorates due to unused hardware standing by
Solution Approach 1:
The patent implements dynamic infrastructure adjustment by using traffic modeling predictions to flexibly allocate storage resources. The system continuously monitors actual traffic patterns and adjusts hardware activation states (active/standby) based on predicted future demand, transforming the fixed infrastructure into a dynamic system that adapts to varying workload requirements while maintaining service quality.
Solution Approach 2:
The system performs preliminary traffic modeling and prediction to forecast future storage demands before they occur. By analyzing historical traffic data and generating predictions in advance, the system can proactively adjust infrastructure configuration and resource allocation ahead of actual demand spikes, preventing service degradation while avoiding unnecessary hardware activation.
2Productivity
If more hardware is added to strengthen infrastructure, then capacity to support increasing client requirements is improved, but cost and complexity increase
Solution Approach 1:
The patent implements self-service through automated traffic modeling and prediction systems that autonomously analyze storage traffic patterns, forecast future demands, and generate optimization recommendations without requiring manual intervention. The system automatically correlates traffic models with infrastructure configurations and identifies optimal resource allocation strategies, enabling the infrastructure to self-manage its capacity and complexity.
Solution Approach 2:
The system establishes feedback loops where actual storage traffic data is continuously collected, compared against predicted values, and used to refine future predictions and adjust infrastructure configurations. This closed-loop feedback mechanism enables the system to learn from actual usage patterns and progressively optimize resource allocation, avoiding unnecessary hardware additions while maintaining adequate capacity.
3Measurement precision
If traffic modeling accuracy is improved through continuous learning, then resource allocation precision is improved, but computational overhead increases
Solution Approach 1:
The patent applies partial action by implementing continuous learning and model adjustment only at appropriate intervals and thresholds rather than constantly. The system monitors traffic patterns and triggers retraining or model updates only when significant changes are detected or at scheduled intervals, avoiding unnecessary computational overhead while maintaining sufficient prediction accuracy for effective resource allocation decisions.
Data Source
AI summary
A method and system for storage traffic modeling in a Software Defined Storage (SDS) is disclosed. The method includes the steps of: collecting observed values of at least one performance parameter in a period of time from a storage node; learning a trend structure of the at least one performance parameter varying with time from the observed values; and providing a predicted value of one performance parameter in a particular point in time in the future. The storage node is operated by SDS software. The trend structure is adjusted based on observed values collected after the period of time. The predicted value is an output of the trend structure which has been adjusted.