Hybrid Cloud Workload Placement via Simulation Forecasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In a hybrid cloud environment, existing workload placement techniques face challenges due to the large number of available cloud options, making it impractical to measure performance across all combinations, and often lack historical performance data, leading to inefficient and costly deployment processes.
Innovation Solution
Deploying a lightweight simulation application across multiple cloud options to forecast performance without actual deployment, using vertical and horizontal relationship functions to predict performance values, allowing for optimized placement without extensive measurement or historical data reliance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If workload placement measures performance across all cloud options, then placement accuracy is improved, but computational cost and time increase significantly
Solution Approach 1:
The patent performs preliminary actions by deploying simulation applications beforehand to collect performance data from multiple cloud options. This pre-collected data is stored and reused for subsequent workload placement decisions, eliminating the need to measure all options at deployment time and thus reducing computational cost and deployment time while maintaining placement accuracy.
2Measurement precision
If workload placement uses actual deployment testing, then performance measurement accuracy is improved, but computational resources and costs increase
Solution Approach 1:
The patent uses simulation applications as copies that mimic the behavior of actual workloads without requiring full deployment. These simulation copies generate performance data that accurately reflects real workload characteristics while consuming minimal computational resources, thus maintaining measurement precision while dramatically reducing computational cost.
3Productivity
If workload placement relies on historical performance data, then deployment speed is improved, but adaptability to new cloud options decreases
Solution Approach 1:
The patent creates a universal performance database that serves multiple purposes: it stores historical data for quick deployment decisions, accommodates new cloud options through continuous data collection, and supports both simulation-based and actual deployment scenarios. This multi-functional system maintains deployment speed while adapting to new cloud environments.
4Manufacturing precision
If the system evaluates all cloud options for workload placement, then placement optimization is improved, but system complexity increases
Solution Approach 1:
The patent segments the workload placement process into distinct phases: data collection through simulation, data storage in a database, and query-based decision making. This segmentation allows the system to evaluate all cloud options comprehensively for optimal placement while managing complexity through modular, reusable components that can be independently maintained and scaled.
Data Source
AI summary
An application is deployed on a first cloud from a set of clouds. A simulator is deployed on each cloud in the set. A vertical relationship function is computed between a time-series of the application and a time-series of the simulator. A first actual value in the time-series of the application is forecasted for a future time. A horizontal relationship function is computed between a first simulator value in the time-series of the simulator on the first cloud and a second simulator value in the time-series of the simulator on a second cloud. A second actual value in the time-series of a hypothetical deployment of the application on the second cloud is forecasted for the future time without deploying the application on the second cloud. The application is placed on the second cloud when the second actual value satisfies a condition.