EventSentry has a unique timer-filter feature that can suppress unnecessary alerts if the error condition only lasts for a (configurable) short period of time. This can be applied to alerts about hosts being down, services being stopped and more.
In this HowTo we will use the example of heartbeat events, which can sometimes generate unnecessary alerts due to a temporary network issue or packet drop that may correct itself automatically.
The diagrams below illustrate the two possible scenarios for filter timers:
Scenario 1: Offline host doesn't come back online within timer period (4 min) and 2 alerts are generated.
Scenario 2: Host comes back online within timer period and no alerts are generated
Under "Event Logs," create a new package labeled "Heartbeat Timer Filter," right-click and assign this to only the management console machine.
Fig.1: Creating and Assigning a New Package
Right-click the package and create an include filter labeled "Heartbeat Down" with the following details:
Fig.2: Filter Heartbeat Down
Right-click the package and create another include filter labeled "Heartbeat Up" with the following details:
Fig.3: Filter Heartbeat Up
Click on the filter "Heartbeat Down", click the "Timers" tab, check "Enable Timer" and set an appropriate timeout, e.g. 4 minutes. Click the '+' under "Filters that clear this timer" and select "Heartbeat Up"; under "Insertion Strings," click the '+' and choose "1-1."
Fig.4: Configuring Timers