Computational Workload Forecasting for Proactive Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud service providers face challenges in predicting and preventing errors, failures, and outages in cloud-based systems due to reactive monitoring techniques, which often fail to address issues before they occur.
Innovation Solution
A system utilizes time-series machine learning models to forecast computational workloads by training and testing multiple versions of time-series models with filtered parameter combinations, selecting the most accurate model, and detecting anomalies by comparing forecasts to threshold values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If reactive monitoring techniques are used to detect system issues, then the monitoring process is simple and straightforward, but errors, failures, and outages occur before remedial action can be taken
Solution Approach 1:
The system performs preliminary actions by training multiple time-series forecasting models on historical workload data and generating forecasts of future computational workloads. These forecasts are produced in advance before actual anomalies occur, enabling the system to predict potential issues and take remedial action proactively rather than reactively, thus improving reliability while maintaining timely response.
2Measurement precision
If multiple versions of time-series models are trained and tested with filtered parameter combinations to improve forecasting accuracy, then prediction accuracy is improved, but the computational complexity and training time increase
Solution Approach 1:
The system segments the model training process by dividing it into distinct phases: (1) generating multiple versions of time-series models with different parameter combinations, (2) filtering parameter combinations based on performance criteria, and (3) selecting the most accurate model for deployment. This segmentation allows the system to manage complexity systematically while achieving high forecasting accuracy through comparative evaluation of multiple model versions.
Solution Approach 2:
The system applies parameter changes by training multiple versions of time-series models with different parameter combinations and using filtered parameter sets to generate diverse model variants. By systematically varying model parameters and evaluating their performance, the system identifies optimal parameter configurations that maximize forecasting accuracy while understanding the trade-offs involved in model complexity.
3Reliability
If forecasting models are used to predict future computational workloads, then proactive anomaly detection is enabled, but the computational resources required for model training and forecasting increase
Solution Approach 1:
The system applies partial action by training and evaluating multiple versions of forecasting models with filtered parameter combinations rather than exhaustively training all possible model variants. This approach achieves sufficient forecasting accuracy for reliable anomaly detection while consuming fewer computational resources than a complete exhaustive search would require, balancing detection capability with resource efficiency.
Data Source
AI summary
Techniques for predicting anomalies in forecasted time-series data are disclosed. A system. A system predicts whether a monitored computing system will experience anomalies by comparing forecasted values associated with components in the monitored computing system to threshold values. The system utilizes time-series machine learning models to forecast workloads of computing resources in the monitored computing system. The system trains and tests multiple different versions of a time-series model and selects the most accurate version to generate forecasts for a particular workload in the computing system. The system compares the forecasts to threshold values to predict anomalies. Based on detecting anomalies, the system generates recommendations for remediating predicted anomalies.