Self-Adjusting Policies via Semi-Supervised Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional management and monitoring applications rely on static policies that become obsolete due to the dynamic nature of virtualization and cloud environments, making it difficult for system administrators to accurately manage and monitor computer infrastructure as workloads and infrastructure constantly change.
Innovation Solution
The use of semi-supervised machine learning to continuously adjust operational thresholds and policies in a computer infrastructure, providing a comprehensive overview of compute, storage, and network attributes, allowing for self-adjustment based on infrastructure and workload transformations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If static policies are used to manage computer infrastructure, then system administrators can define and control performance thresholds, but the policies become obsolete as workloads and infrastructure constantly change
Solution Approach 1:
The patent transforms static management policies into dynamic ones by implementing machine learning models that continuously learn from infrastructure data and automatically adjust thresholds. The system evolves policies over time to adapt to changing workloads and infrastructure conditions, resolving the contradiction between policy reliability and adaptability.
Solution Approach 2:
The system enables self-adjusting policies where the infrastructure management system automatically modifies its own thresholds and parameters based on learned patterns from historical data. This self-service capability eliminates the need for continuous manual policy updates while maintaining effectiveness in dynamic environments.
2Loss of information
If detailed monitoring of all infrastructure attributes is implemented, then comprehensive understanding of system state is achieved, but system complexity and computational overhead increase
Solution Approach 1:
The patent extracts only the most relevant features and attributes from the complete infrastructure data using feature selection techniques. The machine learning model focuses on key performance indicators and critical metrics rather than processing all available data, reducing system complexity while maintaining information completeness for effective decision-making.
Solution Approach 2:
The system applies different monitoring granularities to different infrastructure components based on their importance and variability. Critical resources receive detailed monitoring while less critical resources use aggregated metrics, optimizing the balance between information completeness and system complexity through localized monitoring strategies.
Data Source
AI summary
Embodiments relate to a method for managing and analyzing a computer environment. The method includes receiving, by the host device, a set of data elements from at least one computer environment resource of the computer infrastructure, each data element of the set of data elements relating to an attribute of the at least one computer environment resource. The method includes applying a system analysis function to the set of data elements to characterize a dataset specification associated with the set of data elements. The method includes receiving, by the host device, a user-selected policy threshold criterion based on the dataset specification and providing the user-selected policy threshold criterion to the semi-supervised learning algorithm as a parameter. The method includes adjusting a boundary of the dataset specification of the set of data elements, as associated with the user-selected policy threshold criterion, based on a behavioral change of the computer infrastructure.