Flapper Detection Strategy for Configuration Item Synchronization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In remote network management, updating the service model to account for minor changes, known as 'flapper values,' can be wasteful in terms of memory and processing resources, leading to poor system performance.
Innovation Solution
A flapper detection strategy is implemented to identify flapper values and prevent unnecessary re-computation of the service model by distinguishing between significant and minor changes, using a flapper candidates table and detection strategies to determine whether a new value is a flapper value.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the service model is re-computed for every change in CI records, then the service model remains synchronized with the current state, but processing resources are wasted on minor changes
Solution Approach 1:
The system performs preliminary classification of changes by comparing the magnitude of change against a threshold before initiating re-computation. This preliminary action filters out insignificant changes that would not affect service model accuracy, thereby avoiding wasted processing resources while maintaining synchronization for meaningful changes.
Solution Approach 2:
The system changes the parameter being monitored from absolute change to relative change by introducing a threshold value. Instead of re-computing for any change, the system only triggers re-computation when the change exceeds the threshold, transforming the parameter evaluation criterion to eliminate unnecessary processing.
2Reliability
If the service model is re-computed for every change in CI records, then the service model remains up-to-date, but system performance deteriorates
Solution Approach 1:
The system performs preliminary classification of changes by comparing the magnitude of change against a threshold before initiating re-computation. This preliminary action filters out insignificant changes that would not affect service model accuracy, thereby avoiding wasted processing resources while maintaining synchronization for meaningful changes.
Solution Approach 2:
The system changes the parameter being monitored from absolute change to relative change by introducing a threshold value. Instead of re-computing for any change, the system only triggers re-computation when the change exceeds the threshold, transforming the parameter evaluation criterion to eliminate unnecessary processing.
3Device complexity
If all changes are treated equally, then the system is simple to implement, but memory and processing resources are wasted
Solution Approach 1:
The system changes the parameter being monitored from absolute change to relative change by introducing a threshold value. Instead of re-computing for any change, the system only triggers re-computation when the change exceeds the threshold, transforming the parameter evaluation criterion to eliminate unnecessary processing.
Solution Approach 2:
The system applies different quality treatment to different changes based on their magnitude. Significant changes trigger full re-computation while minor changes are filtered out. This local differentiation in processing quality optimizes resource usage without requiring complex system architecture.
Data Source
AI summary
A computational instance of a remote network management platform includes a persistent storage that contains a plurality of CI records corresponding to a set of computing devices disposed within a managed network, a set of software applications configured to execute on the set of computing devices, and a network-based service that is provided by execution of the set of software applications. The computational instance also includes one or more server devices configured to: receive an indication of a change to a CI record of the plurality of CI records; add, to a flapper candidates table in the persistent storage, a new value for a field of the CI record; determine, using a flapper detection strategy, flapper strategy data based on the new value and at least one old value for the field; and determine, based on the flapper strategy data, whether to trigger a re-computation of the service model.