Hi!

is there a way to receive down notifications per device instead of per sensor? If let's say a device has 8 sensors and the whole device gets down, I would like to have only 1 down notification instead of 8. The same goes with up messages.

I was trying to recreate this behavior with notification summarizations, but I was unable to do that since summarizations affect every sensor on every device plus I receive a summarization email after a selected number of minutes (so at least 2 notifications instead of one).

In short: Is there a way to receive only one notification per device, no matter how many sensors are down? Until everything is up again, of course.

Article Comments

Dear mq

If you create a device via an auto-discovery device group, the ping sensor for each device will be set to "Master object for parent". This works this way:

  • The parent of the sensor is the device
  • If the ping is down, that sensor acts as dependency 'master' for the device
  • Meaning, if the ping is down, it pauses the device, and with it, all other sensors
  • Once the ping gets up again, the device will be resumed

You can manually set this option via the "Settings" tab of a ping sensor. This option is not available for multi-edit, meaning you have to go through each ping sensor manually. If you have devices without a ping sensor, please create it.


Jun, 2017 - Permalink

Hi,

thanks for the answer!

I have already found this feature, but... I can only set one sensor as "Master object of parent", right? I know from experience that ping is often the last sensor to get down, so I would still get multiple down notifications from every other sensor if ping is still up, wouldn't I?

Is there a way to pause every other sensor of a device regardless which sensor gets down first?


Jun, 2017 - Permalink

Dear mq

The solutions is to set the ping sensor to a 30 second scanning interval, and the notification on a 60 second delay before it triggers. Also, the ping should be set to get instantly down (instead to warning first, this option is available in the same section where you set the scanning interval.)


Jun, 2017 - Permalink

Hi. Thanks again for the fast answer.

I didn't mean "ping is often the last sensor to get down" literally. What I meant was that ping often stays up even if other sensors on a device went down.

For example: We had a problem with a Server. The Volume Service went down, the Server Service went down etc. but Ping stood up forever. After a restart everything was fine but until then we had a couple of notifications including some notification summaries.

In this scenario - no matter how I change the scanning interval - setting ping as the master object for parent wouldn't help, would it?

I changed the notification summarization time which helped a bit. And I will set ping as the master object for parent which will help if ping gets down. But is possible to stop notifications of a device if ANY sensor of this device gets down?

I hope you understand what I mean.

Thanks!


Jun, 2017 - Permalink

Dear mg

In this case, it should be investigated why the ping is not going down. Is it because the device itself is still up and just some services, or sensors fail? Or is the device actually down? In that case, please check the "Logs" tab of that ping sensor. Do you see any status change there?


Jun, 2017 - Permalink

Hi,

the device (a vm) was still up and technically running, but not working and we couldn't connect via rdp. After a reboot everything worked fine again.

Unfortunately there are no ping logs for the moment the problem happened.


Jun, 2017 - Permalink

Dear mg

Is there any other sensor type which reliably goes into the alarm status when the device is down? The "master object for parent" is available for all sensor types, but only one sensor per device can have this enabled. (If you enable it for another sensor, this property will be removed from the previous sensor.)


Jun, 2017 - Permalink