MENU
An alert has been created to email when AppInsight for SQL detects a database fragmentation that exceeds 30%. The issue is not considered resolved unless the fragmentation is below 10%. How would the reset conditions be configured for this alert?
In advanced alerting scenarios, particularly with performance metrics like disk fragmentation or temperature, the 'Trigger Condition' and 'Reset Condition' often require different thresholds to prevent 'flapping'---a state where an alert rapidly toggles between triggered and reset states because the metric is hovering right at the threshold line. According to the SolarWinds Platform Alerting Guide, while the default behavior is to reset when trigger conditions are no longer true (Option D), this would reset the alert as soon as fragmentation hit 29.9%.
To satisfy the specific requirement where the issue is only 'resolved' at 10%, a custom Reset Condition must be defined. By selecting 'set reset condition to reset when condition is below a set percent,' the administrator can explicitly define a separate value ($10%$) from the trigger value ($30%$). This creates a 'deadband' or hysteresis effect, ensuring the alert remains active and visible until the database maintenance has successfully reduced the fragmentation to the desired healthy level. This configuration is essential for AppInsight applications where returning to a 'not-critical' state (e.g., 29%) does not necessarily mean the underlying performance bottleneck has been sufficiently remediated.