Resource-Aware Data Mining for IT Infrastructure
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern IT infrastructures face challenges in managing and analyzing vast amounts of operational data from complex technical environments, leading to increased overhead and inefficiencies in data collection, as existing methods fail to differentiate between normal and abnormal data, resulting in resource wastage and suboptimal system health maintenance.
Innovation Solution
A resource-aware dynamic operational data collection and analysis plan is generated and implemented, using real-time and historical data to evaluate health and resource states, identify abnormal components, and determine data collection strategies based on topology, thereby streamlining data collection and focusing on relevant data for problem prediction and system health.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing tools collect and analyze all operational data from the full stack to gain intelligence about the technical environment, then measurement precision and reliability are improved, but device complexity and loss of energy increase due to the large quantity of data produced, collected, stored, and analyzed
Solution Approach 1:
The patent segments the operational data collection process by identifying and prioritizing critical objects and parameters based on system health importance. Instead of uniformly collecting data from all objects, the system divides data collection into targeted segments focused on abnormal or critical objects, reducing the overall data volume while maintaining essential visibility.
Solution Approach 2:
The patent applies local quality by adjusting data collection intensity and granularity based on the specific characteristics and importance of individual objects. Critical objects receive higher monitoring priority with more detailed data collection, while non-critical objects receive reduced monitoring, optimizing resource allocation across the technical environment.
2Reliability
If existing tools collect and analyze all operational data to gain intelligence, then reliability of system monitoring is improved, but loss of time and productivity decrease due to the overhead challenge presented by the quantity of data
Solution Approach 1:
The patent performs preliminary action by pre-identifying critical objects and their associated parameters before data collection begins. The system establishes priority lists of objects that require monitoring based on their importance to system health, enabling proactive filtering of data collection targets and reducing unnecessary data processing.
Solution Approach 2:
The patent applies partial action by collecting data only from objects that are abnormal or critical, rather than from all objects uniformly. This selective approach collects sufficient data to maintain system health monitoring reliability while avoiding the excessive data collection that would burden processing resources.
3Measurement precision
If tools collect data from the full stack including hardware, OS, virtualization layer, middleware, applications, and third party services, then measurement precision improves, but device complexity and loss of substance increase due to the large volume of data requiring management
Solution Approach 1:
The patent extracts and isolates only the critical data elements needed for system health monitoring from the full operational data set. By identifying and extracting data from abnormal or critical objects across the technical stack, the system removes unnecessary data from storage and processing pipelines, reducing overhead while preserving essential monitoring capabilities.
Data Source
AI summary
Computer implemented method, systems, and computer program products include program code executing on a processor(s) monitors objects in the technical environment to collect real-time operational data. The program code obtains historical operational data and historical resource data from the technical environment and generates models to evaluate health states and resource states of the one or more objects. The program code applies the models and identifies objects as abnormal. The program code obtains a topology of the technical environment to identify objects impacted by the abnormal object(s). The program code generates a resource aware dynamic operational data collection plan.