Machine Learning Capacity Reservations for Cloud Application Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual capacity reservations for cloud-based computing instances are prone to human errors, resulting in mismatches and inefficiencies, particularly during application failover, which can lead to instance unavailability and increased costs.
Innovation Solution
A capacity reservation machine learning (ML) model uses linear regression techniques to analyze current fleet trends and automate capacity reservations, enhancing accuracy and reducing errors by aligning reservations with actual needs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual capacity reservations are used, then ease of operation is improved, but reliability deteriorates due to human errors and mismatches
Solution Approach 1:
The system performs self-service by automatically analyzing fleet data and creating capacity reservations without manual intervention. The machine learning model processes historical data, predicts future capacity needs, and executes reservations autonomously, eliminating human errors while maintaining operational simplicity through automated decision-making
Solution Approach 2:
The patent replaces the mechanical manual reservation process with an automated machine learning system. The ML model substitutes human judgment and manual operations with algorithmic analysis, using historical data patterns to automatically determine and execute capacity reservations, thereby improving reliability while reducing operational complexity
2Device complexity
If manual capacity reservations are used, then device complexity is reduced, but productivity deteriorates due to laborious tasks and time consumption
Solution Approach 1:
The system achieves self-service automation where the machine learning model independently analyzes fleet data, predicts capacity requirements, and executes reservations without manual intervention. This autonomous operation dramatically increases productivity by eliminating the laborious manual process while the system manages its own complexity through integrated data processing and algorithmic decision-making
Solution Approach 2:
The patent applies parameter changes by transforming the reservation process from manual to automated through machine learning algorithms. The system changes the fundamental parameters of how reservations are made - from human-based manual creation to algorithm-based automated prediction and execution, thereby increasing productivity while the complexity is managed through systematic data processing
3Ease of operation
If ad-hoc capacity reservations are used, then ease of operation is improved, but reliability deteriorates leading to instance unavailability during failover
Solution Approach 1:
The system performs preliminary action by proactively analyzing historical fleet data and predicting future capacity needs before failures occur. The machine learning model continuously monitors patterns and automatically creates capacity reservations in advance, ensuring instances are pre-positioned and available when failover events happen, thereby improving reliability while maintaining operational simplicity
Solution Approach 2:
The patent implements feedback mechanisms where the machine learning model continuously analyzes actual fleet usage patterns and adjusts future reservation predictions accordingly. This feedback loop ensures the system learns from historical performance and optimizes its reservation strategies, improving reliability by adapting to changing conditions while maintaining ease of operation through automated continuous improvement
Data Source
AI summary
Embodiments disclosed are directed to a computing system that performs operations for leveraging machine learning to automate capacity reservations for application failover in a cloud-based computing system. The computing system determines a simulated usage capacity of a set of applications executing in a first zone of a cloud-based computing system. The computing system then determines an amount of cloud-based computing instances in a second zone of the cloud-based computing system needed to maintain the simulated usage capacity in an event of a failover of the first zone. Subsequently, the computing system reserves the amount of cloud-based computing instances in the second zone.