Real-Time Network Flaw Detection via Metric Stream Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network monitoring systems face delays in detecting operational flaws due to reliance on database queries, which can take time and may not provide real-time performance metric analysis.
Innovation Solution
A computer-implemented method where a network management server generates a stream of tuples containing network element identifiers, performance metrics, and object instance identifiers, allowing a monitoring server to detect operational flaws in real time without accessing a database, using control metrics as a heartbeat to set flags for alert generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If database queries are used to collect performance metrics, then comprehensive network performance data can be obtained, but detection delay increases and real-time monitoring is not achieved
Solution Approach 1:
The patent extracts the essential monitoring function from the database query system by implementing a stream processing approach that selectively processes only relevant performance metrics in real-time, rather than querying entire databases. This extraction of necessary data elements enables timely detection without comprehensive database access.
Solution Approach 2:
The system performs preliminary actions by pre-processing and filtering performance metrics as they arrive in the stream, preparing data for immediate analysis before detection is needed. Control metrics are used to pre-establish baseline conditions, enabling faster real-time flaw detection without waiting for database queries to complete.
2Speed
If real-time monitoring of performance metrics is implemented, then operational flaws can be detected quickly, but system complexity increases
Solution Approach 1:
The monitoring system is segmented into distinct functional components: a stream receiving module that ingests performance metrics, a processing module that applies detection logic, and an alerting module that notifies operators. This segmentation allows real-time processing while keeping each component's complexity manageable and well-defined.
Solution Approach 2:
The patent introduces control metrics as an intermediary mechanism that mediates between raw performance metrics and flaw detection logic. These control metrics serve as reference values that simplify the detection process by providing pre-established benchmarks, reducing the computational complexity required for real-time analysis.
3Reliability
If continuous monitoring of all performance metrics is performed, then comprehensive operational flaw detection is achieved, but processing overhead and resource consumption increase
Solution Approach 1:
The system applies partial action by monitoring only the specific performance metrics that are relevant to operational flaw detection, rather than continuously processing all available metrics. The stream processing approach selectively applies detection logic to pertinent data elements, reducing resource consumption while maintaining reliable flaw detection for critical parameters.
Data Source
AI summary
In an embodiment, a computer-implemented method receives and monitors performance metrics from network element. The method also includes receiving periodic control metrics corresponding to object instances. Performance metrics and control metrics provide information about operation of object instances. By monitoring the metrics, a network server is able to detect an operational flaw in the network. Monitoring the performance and control metrics in real time increases the speed of detecting any operational flaw in the network.