Application Transaction Analysis Using Historical Averages
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing monitoring systems rely on CPU utilization metrics, which may not accurately reflect the functional performance of computer systems, potentially masking issues that affect customer service, such as increased response times or error rates in application transactions.
Innovation Solution
A system and method for network application transaction analysis that measures application transaction metrics like response time and success rates, calculates historical averages, and outputs alarms or notifications based on threshold differences, enabling early detection of performance issues specific to applications rather than just servers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If CPU utilization metrics are used for monitoring, then server performance can be tracked, but application-specific performance issues may be masked
Solution Approach 1:
The patent segments the monitoring system into two distinct metric collection paths: one for server-level metrics (CPU utilization) and another for application-level metrics (transaction response times, error rates). This segmentation allows each layer to be monitored independently with appropriate metrics, preventing the loss of application-specific performance information while maintaining server performance tracking capability.
Solution Approach 2:
The patent adds a new dimension to performance monitoring by introducing application transaction metrics alongside traditional server metrics. Instead of relying solely on CPU utilization (single dimension), the system now monitors multiple dimensions including transaction response times, error rates, and throughput specific to each application, enabling precise detection of application performance issues.
2Reliability
If application transaction metrics are monitored, then application performance issues can be detected early, but monitoring complexity increases
Solution Approach 1:
The patent implements a universal monitoring framework that can handle both server metrics and application metrics through a single system architecture. The monitoring tool is designed to collect, store, and analyze multiple types of metrics (CPU utilization, transaction response times, error rates) using common infrastructure, thereby reducing overall system complexity despite the increased number of monitored parameters.
Solution Approach 2:
The patent introduces an intermediary monitoring tool that acts as a mediator between the application/server and the analysis system. This intermediary component captures application transaction metrics, normalizes them, and presents them in a unified format, simplifying the monitoring architecture and reducing the complexity of integrating multiple monitoring sources.
3Measurement precision
If historical averages and threshold comparisons are implemented, then false alarms can be reduced, but calculation and processing time increases
Solution Approach 1:
The patent implements preliminary action by pre-calculating and storing historical average metrics and threshold values in a database. Instead of computing these values in real-time during alarm evaluation, the system prepares the baseline data in advance, allowing rapid comparison against current metrics and significantly reducing processing time for alarm generation.
Solution Approach 2:
The patent replaces complex real-time statistical analysis with simpler threshold-based comparison mechanics. By substituting sophisticated continuous analysis with pre-defined threshold checks against historical averages, the system achieves high alarm accuracy while minimizing processing time and computational overhead.
Data Source
AI summary
A system provides network application transaction analysis. An analysis tool, when executed by a processor, measures an application transaction metric and calculates a historical system average and a historical application average associated with the application transaction metric based on server performance logs. The tool determines whether the application transaction metric differs from the historical system average by more than a first threshold amount during consecutive measurements. The tool determines whether the application transaction metric differs from the historical application average by more than a second threshold amount during consecutive measurements if the application transaction metric does not differ from the historical system average by more than the first threshold amount during consecutive measurements. The tool outputs an alarm to a user interface to enable a generation of an investigation if the application transaction metric differs from the historical application average by more than the second threshold amount during consecutive measurements.