Server Performance Monitoring via Weighted Activity Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large server farms face inefficiencies due to idle servers consuming significant power and resources, with existing monitoring methods failing to effectively identify non-useful server activities, leading to wastage and space utilization issues.
Innovation Solution
A computer-implemented method that monitors server performance by determining activity metrics, calculating weighted combinations of these metrics to provide a certainty value indicating whether a server is serving a useful purpose, allowing for the exclusion of non-useful activities and enabling informed decision-making for power control and resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If servers are kept running to ensure availability, then service reliability is improved, but power consumption increases significantly
Solution Approach 1:
The system dynamically adjusts server power states based on real-time performance monitoring. Servers transition between active, idle, and powered-off states according to their actual workload and performance metrics, rather than remaining statically powered on. This dynamic state adjustment resolves the contradiction by maintaining reliability only when necessary while reducing power consumption during low-utilization periods.
Solution Approach 2:
The invention changes the operational parameters of servers by introducing performance thresholds and weighted performance scores. When performance metrics fall below certain thresholds or scores indicate low usefulness, servers are transitioned to lower power states. This parameter-based control mechanism enables automatic adaptation between reliability and power consumption based on actual server utility.
2Ease of manufacture
If traditional monitoring methods are used to track server activity, then implementation simplicity is maintained, but accuracy in identifying non-useful servers deteriorates
Solution Approach 1:
The invention introduces an intermediary performance evaluation layer that sits between traditional monitoring and decision-making. Instead of directly using raw monitoring data, the system calculates weighted performance scores that combine multiple metrics (CPU usage, memory usage, I/O activity, network traffic) with configurable weights. This intermediary calculation layer improves detection accuracy while maintaining implementation simplicity through modular, configurable weight parameters.
Solution Approach 2:
The performance scoring mechanism serves multiple functions simultaneously: it identifies non-useful servers, ranks servers for consolidation candidates, triggers power state transitions, and provides visibility for management. This multi-functional approach improves measurement precision across different use cases while maintaining a single, simple implementation framework.
3Measurement precision
If multiple performance metrics are monitored to accurately assess server usefulness, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The system extracts and focuses on the essential performance metrics that truly indicate server usefulness, rather than monitoring all possible system parameters. By selecting key metrics (CPU, memory, I/O, network) and applying weighted scoring, the invention achieves high measurement precision while avoiding the complexity of comprehensive multi-parameter monitoring. The weighted approach extracts the most significant signals from the data.
Solution Approach 2:
The invention merges multiple performance metrics into a single weighted performance score. Instead of managing separate thresholds and alert systems for each metric, the system combines CPU usage, memory usage, I/O activity, and network traffic into one composite score that drives decision-making. This merging reduces system complexity while maintaining comprehensive assessment accuracy.
Data Source
AI summary
A computer implemented method monitors a computer to determine values of a plurality of activity metrics of the monitored computer. A weighted combination of functions of the determined values is calculated as a measure of performance of the monitored computer. The weighted combination may be a weighted combination of net values of the activity metrics. In one method, the net values are calculated as the determined values of the plurality of activity metrics of the monitored computer excluding contributions to the values from one or more predetermined activities. In another method, the net values are calculated as total values of the respective activity metrics from which the total values of the one or more predetermined activities are subtracted. The weighted combination may be used to control power consumption or otherwise take action in relation to the computer.