Data Center Asset Telemetry Interpolation During Network Lapses
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data center asset telemetry data can be incomplete or missing due to network connectivity lapses, leading to negative user experiences and inaccurate predictive analytics or billing calculations.
Innovation Solution
A method involving receiving historical data center asset telemetry data, training a prediction model, and interpolating additional data using the generated predictions to manage and monitor data center assets effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If network connectivity is used to collect data center asset telemetry data, then data collection capability is improved, but data completeness deteriorates due to network connectivity lapses
Solution Approach 1:
The system performs preliminary actions by collecting and storing telemetry data during network connectivity periods, then uses historical data to generate predictions that fill in missing data points during connectivity lapses. This ensures continuous data availability without requiring constant network connection.
Solution Approach 2:
A prediction model acts as an intermediary between the telemetry collection system and the data processing system. The model receives historical telemetry data and network connectivity information, then generates predicted telemetry values that bridge gaps caused by network interruptions, ensuring data completeness.
2Reliability
If historical data is used to train prediction models, then data completeness is improved through interpolation, but computational complexity increases
Solution Approach 1:
The system changes the approach from collecting raw telemetry data continuously to collecting data at strategic time points and using predictions to fill gaps. This parameter change in data collection frequency combined with prediction-based interpolation reduces computational complexity while maintaining data completeness.
Solution Approach 2:
Instead of continuously monitoring all asset parameters in real-time, the system creates a simplified representation by training prediction models on historical data patterns. These models copy the essential behavior patterns and generate predictions that replicate actual telemetry values without requiring continuous complex processing.
3Speed
If real-time data collection is performed, then data freshness is improved, but data accuracy deteriorates due to network connectivity issues
Solution Approach 1:
The system uses feedback from historical data to improve future data collection and prediction accuracy. By analyzing past telemetry data and network connectivity patterns, the system refines prediction models that can accurately interpolate missing values, ensuring both freshness and accuracy of the complete data set.
Solution Approach 2:
The system performs preliminary data collection during network connectivity periods and uses this historical data to train prediction models. These models then generate accurate predictions for periods with connectivity lapses, maintaining data accuracy without requiring continuous real-time collection.
Data Source
AI summary
A system, method, and computer-readable medium for performing a data center monitoring and management operation. The data center monitoring and management operation includes: receiving data center asset data, the data center asset data comprising historical data center asset telemetry data; training a prediction model to generate predictions using the historical data center asset telemetry data; and, interpolating additional data center asset telemetry data using the predications generated using the prediction model.