We have a PRTG Cluster in two different network segments since 3 years now and it works pretty good. Sometimes the failover cluster node sets a sensor to warning/failure status (maybe because of latency) with Code: PE015 (could not connect), although everything is fine. The other node states the sensor still as OK.
Node1: Warnung Verbindung konnte nicht hergestellt werden (Code: PE015) Node2: Ok OK
But the Admin gets, of course, a notification because of the notification trigger.
How can we achive that PRTG send a notification only when both nodes have the same status, e.g both are in state warning? Is this possible? Thanks!
Article Comments
Thanks for your reply. I understand the conditions for Down/Partial Down. The Problem is, when one Cluster Node states a WARNING (because of any strange issue) and the other Cluster Node states OK (which is the real Sensor status), we get a notification of Warning/OK. But the notification should be only send when BOTH Nodes state WARNING.
Jan, 2017 - Permalink
Hi there,
As requested before, what state triggers (or other triggers) are currently defined for the affected sensor/device/group?
Jan, 2017 - Permalink
Two Triggers:
Trigger Zustand: Wenn der Sensor für mindestens 180 Sekunden im Zustand WARNUNG ist, führe E-Mail an Administrator aus. Wenn der Sensor für mindestens 300 Sekunden im Zustand Warnung ist, führe keine Benachrichtigung aus und wiederhole dies alle 0 Minuten Wenn die Bedingung nicht mehr zutrifft, nachdem eine Benachrichtigung ausgelöst wurde, führe E-Mail an Administrator aus.
Trigger Zustand: Wenn der Sensor für mindestens 180 Sekunden im Zustand FEHLER ist, führe E-Mail an Administrator aus. Wenn der Sensor für mindestens 300 Sekunden im Zustand Fehler ist, führe keine Benachrichtigung aus und wiederhole dies alle 0 Minuten Wenn die Bedingung nicht mehr zutrifft, nachdem eine Benachrichtigung ausgelöst wurde, führe E-Mail an Administrator aus.
Jan, 2017 - Permalink
Hi,
So the reason why you are receiving a "WARNING" notification (even when just one node is complaining) is, that there is no "Partial Warning" status. If a sensor is down on just one node you will have a sensor in the "Partial Down" status and you will just receive a notification as soon as the sensors changes to "Down" (when the sensor is down on all Nodes).
If sensor "Partial Down" (Down on one Node) > No Notification
If sensor "Down" (Down on all Nodes) > Notification
If sensor "Warning" (no matter on how many nodes) > Notification
Jan, 2017 - Permalink
Hi there,
Actually, if you created a state trigger for the "Down" condition, it will only be triggered if all Cluster Nodes are reporting the Down status as the sensor would be otherwise only Partial Down. If you created a state trigger for Partial Down, you will receive a notification just when one of the Cluster Nodes is reporting an error.
What are you current defined triggers?
Jan, 2017 - Permalink