Hi All,
I have been tuning some alert definitions for a client I am working with at the moment who have a new Orion installation with NPM 11.5. The node down alert has been duplicated and tuned to exclude a specific set of nodes which match a custom property value in line with the client's requirements. The trigger condition was set correctly, but the wrong value was specified in the reset condition, resulting in triggered alerts not resetting. The alert definition has since been corrected, but the already triggered alerts do not reset, even though the reset condition would now be met.
The alert definition summary confirms that 0 nodes will trigger the alert, hence the 9 triggered alerts in theory would then reset after 1 minute once the alerts are checked by the alerting engine. I will clear the alert manually, but wanted to check with others that they have also experienced this situation and if this is a known issue, or expected behaviour. Historically with the Alert manager GUI in previous releases of NPM (yes, going back a bit now), any changes to alert definitions resulted in prompt trigger/reset actions on all objects.
This observation also exists for down groups where a dependency was not configured and now a parent/child dependency has been defined for 2 groups, making parent down, child unreachable, but the child group still has a group down alert trigger condition and has not reset the alert.
