Implausible data with disconencted probe

If these sensors are part of a dependency tree, then when the remote probe goes down, you can have them inherit a paused state. You can read more about dependencies here.

Dependencies

Once all of the sensors are in a paused state, they will not report any false data or any notifications.

I have a similar problem, similar circumstances. VPN rebuilds on a schedule and this causes a moments loss of connection from Probe to PRTG. During this time 20-30 sensors all pop red, 0GB, 0 processes running, whatever it may be. Notifications are then intermittently fired off and this is troublesome at 2AM in the morning when someone gets called on it.

13.1.2.1463+

Randolfini, you could also set up dependencies like I mentioned before but in your case, I think the best option would be to set up schedules for when your VPN is being rebuilt. This can be done for a group of sensors through their parent device's settings under Schedules, Dependencies, and Maintenance Window.

Group Settings

See the section about Schedules, Dependencies, and Maintenance Window.

Hello,

thanks for your reply!

I have had a similar case again yesterday and I think the dependencies are not the problem. In my case schedules do not apply, since there are no planned VPN renegotiations.

Just to clarify, I have grouped the sensors like this:

Probe -> Group -> (Group ->) Device -> Sensor

For each device I have a ping sensor which is the "master" for the device. What actually went wrong is that the ping sensors still showed green, thus the remaining sensors did not go into the paused or unknown state. And for whatever reason reported data - false data.

As with Randolfini this is quite undesirable for us, since it leads to confusion.

In this case I would have to change all dependencies of all sensors to the probe itself somehow, not overly desirable.

I would like to make a sort of "feature request" out of this, because I think it would be a desirable behavior (and in my opinion the only "correct" behavior) that when a probe disconnects, all sensors of this probe are put into the "unknown" state at once and not after a delay or a refresh period or so.

If the remote probe goes down and there is still data being reported, then something is wrong, they should go into an unknown state. If you could submit a ticket to [email protected] and include your logs for the remote probes that are being affected as well as the core probe, we can look into this.

See here for submitting logs:

https://www.paessler.com/knowledgebase/en/topic/7983-how-can-i-upload-log-files-to-paessler-support-team

Just want to tag that this has been resolved with the latest releases. At least for me.