Database Subsystem Monitoring via Vector Table Scanning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional database monitoring tools consume significant processing resources, leading to performance degradation and inadequate tracking of changes in database subsystems, resulting in ineffective monitoring of subsystem performance.
Innovation Solution
A method utilizing statistical models and subsystem simulations to scan subsystem vector tables, compile vector lists, generate tasks, retrieve logs, determine deviations from average operating states, and provide real-time monitoring with notifications and future scenario predictions, while updating buffer pool distribution settings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional monitoring tools are used to track database subsystems, then monitoring coverage is provided, but processing resources are consumed excessively leading to performance degradation
Solution Approach 1:
The patent introduces an intermediary monitoring system that collects subsystem metrics through lightweight agents rather than directly instrumenting the database. This intermediary layer processes and analyzes metrics externally, reducing the monitoring overhead on the database itself while maintaining comprehensive monitoring coverage.
Solution Approach 2:
The patent creates simplified copies or representations of subsystem states through metric collection and statistical modeling. Instead of monitoring every detail of the actual database subsystems, it monitors copies of their performance characteristics, reducing resource consumption while preserving essential monitoring capabilities.
2Loss of information
If conventional monitoring tools monitor high level metrics, then some performance information is available, but changes in database subsystems are not adequately tracked
Solution Approach 1:
The patent segments the monitoring approach by dividing database subsystems into distinct components (buffer pool, sort heap, statement cache, etc.) and monitoring each subsystem individually. This segmentation enables precise tracking of changes in specific subsystems while maintaining an overall view of database performance.
Solution Approach 2:
The patent implements dynamic monitoring that adapts to changing database conditions. The system continuously updates baseline metrics and adjusts monitoring parameters based on observed patterns, enabling it to detect subtle subsystem changes while filtering out normal variations.
Data Source
AI summary
A method for facilitating real-time monitoring of changes in a database is provided. The method includes scanning a subsystem vector table for a database subsystem that is active on a logical partition; compiling a vector list that includes the active database subsystem; and dynamically generating, based on the vector list, a task for each of the active database subsystem. For each of the active database subsystem, the method further includes retrieving, from a module, a log that includes a parameter by executing the task; determining, by using a model, a deviation from an average operating state of the parameter based on the log; and determining whether the deviation meets a predetermined threshold for a predetermined period of time.


