Capacity-Forecast Model for Data Center Resource Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data center customers face challenges in planning capacity requirements to meet changing demand for application programs, as existing data center management products do not provide adequate tools to calculate the relationship between changes in application program usage and data center resource capacity, leading to reliance on historical data for estimating CPU and memory utilization.
Innovation Solution
A capacity-forecast model is developed using historical capacity and business metric data to estimate resource requirements, allowing for 'what-if' scenario planning and resource reallocation to meet projected business expectations, thereby improving resource utilization and avoiding overprovisioning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data center customers use historical business metric data to estimate capacity changes, then they can make predictions about CPU and memory utilization, but they lack accurate calculation of the relationship between application program usage changes and data center resource capacity
Solution Approach 1:
The patent transforms the capacity planning approach by changing the parameters from manual historical data analysis to an automated model that uses multiple input parameters (business metrics, capacity metrics, correlation coefficients) to generate predictions. The system calculates capacity-forecast models by correlating normalized capacity and business metric data sets, automatically determining relationships between application usage and resource capacity without manual intervention.
Solution Approach 2:
The patent introduces a capacity-forecast model as an intermediary between business metric data and capacity planning decisions. This model acts as a mediator that processes historical data, calculates correlations, and provides automated predictions about capacity requirements, eliminating the need for customers to manually analyze historical data and estimate relationships between usage changes and capacity requirements.
2Adaptability or versatility
If data center management products provide what-if scenario recommendations, then customers can see what can be reclaimed and fitted into current environment, but they cannot plan capacity requirements to meet potential changes in demand for application programs
Solution Approach 1:
The patent makes the capacity planning system dynamic by enabling customers to input different business scenarios and receive updated capacity forecasts. The system calculates capacity-forecast models that can be applied to various what-if scenarios, allowing customers to plan for potential changes in demand by correlating normalized capacity and business metric data sets under different conditions, rather than providing static recommendations.
3Reliability
If customers purchase sufficient computer systems to handle peak demands, then they can meet maximum computational-bandwidth and data-storage needs, but they cannot optimize resource allocation based on actual usage patterns
Solution Approach 1:
The patent implements feedback by using historical capacity and business metric data to calculate correlation coefficients and develop capacity-forecast models. The system continuously analyzes the relationship between application program usage and data center resource utilization, providing feedback that enables customers to optimize resource allocation based on actual usage patterns rather than relying on peak demand estimates, thus reducing overprovisioning while maintaining reliability.
Data Source
AI summary
Methods determine a capacity-forecast model based on historical capacity metric data and historical business metric data. The capacity-forecast model may be to estimate capacity requirements with respect to changes in demand for the data center customer's application program. The capacity-forecast model provides an analytical “what-if” approach to reallocating data center resources in order to satisfy projected business level expectations of a data center customer and calculate estimated capacities for different business scenarios.