I'm trying to setup a threshold trigger for an uptime channel on a device. Easy enough to do. Beyond that, I want to eliminate a redundant notification trigger that will always happen when a manual process occurs as a result of that notification.
I have three identical devices that I monitor. When the device loses power from a power outage, brown out, or other 'building wide' power loss, I want to send out a notification to a group of users so they can power down the device(s) in question, then bring them back up, as this would prevent issues with that device when they ultimately go to use it a week later.
I can think of 2 ways to do this, but I don't believe either way is possible within PRTG currently, so I'm asking if someone knows a way to do it. (Perhaps through custom sensors?)
Possible option 1 - Conditional Formatting. Have an IF statement on the trigger, IF notification sent within 24 hours, do not send another notification.
Possible option 2 - Someway somehow do multi-channel conditional triggers. I.E. If in Group A - Device 1/2/3, system uptime channels >900 seconds, DO NOTIFICATION.
I know it's possible to have inheritance of notifications to make PRTG easier to manage, but I'm not aware of anyway to achieve what I'm trying to do from what research I've done on it so far.
Article Comments
Florian,
Thank you for your detailed response. I'm not sure that any of it applies to what I am trying to achieve though. I'll try to further simplify.
Device A - Uptime Sensor |
Device B - Uptime Sensor |
Device C - Uptime Sensor
All 3 of these devices are located in the same building. I want to have a single alert go out to a mailing list in the event that all 3 devices are >300 Seconds on their uptime sensor.
When PRTG sends that notification out, I will instruct the users to properly power cycle all 3 devices. When they do that, the threshold trigger, regardless of inheritance, will go off again. How do I prevent that second alert, which would effectively be a 'false positive'.
I can of course just tell the users to ignore the alert they create, but I was hoping there was some way with conditional formatting (an if statement) or some sort of multi-device notification. I.E. Custom Uptime Sensor that looks to device A, B & C's Uptime Sensors and only enters a down state when all 3 match the threshold.
Hopefully that clarifies things some. Thank you again for your time.
May, 2022 - Permalink
Hello,
Thank you very much for the clarification!
I do see what you would like to achieve and this should be possible via the usage of sensors limits and with the Business Process sensor. To do so, I invite you to follow the steps below:
- For each Uptime sensor, define an error limit in the channel settings to 300
- Create a new device within the device tree above them
- Add a Business Process sensor to it with the following configuration:
- Channel name: <what you desire>
- Error: 80
- Warning: 50
- Objects: Device uptime sensors
At this point, you should have a configuration similar to this one (without the additional sensors Green IT and OS of the devices):
- Finally, configure the notification trigger to the Busines Process sensor only (example below)
Note: To avoid getting multiple emails, you need to make sure that the notification inheritance is disabled for the devices and their sensors. Only the Business Process sensor should have one.
Hopefully, this will provide the behavior you are looking for. If not, let me know so we can see what would need to be modified.
Regards.
May, 2022 - Permalink
Florian,
That's it! I have not yet been able to test the behavior to ensure it works as intended, but looking over everything I believe I can fine tune it to where this will serve my intended purpose.
While it would still in theory be possible to manufacture a 'faux' down state if all 3 devices are rebooted right on top of each other, it's unlikely and should be a rare occurrence at best. The only time that would likely happen is during maintenance window(s), which that is easily resolved.
Thank you very much for your time and assistance! It is greatly appriceated.
Best Regards, Chris
May, 2022 - Permalink
With pleasure! Glad to hear this is what you are looking for.
Have a great day.
Best regards, Florian
May, 2022 - Permalink
Hello,
Thank you for your message.
To be honest, I'm not quite sure to understand what you desire to achieve exactly. May I have further information about it.
According to your message, you would like to apply a notification trigger to the three devices mentioned and avoid getting multiple notifications from them, is it correct?
If so, I invite you first to make sure that their Ping sensor is defined as master (within the Ping sensor Settings tab, under "Schedules, Dependencies, and Maintenance Window"), this way all their sensors (except Ping) will be paused by dependency and you would receive a notification for the Ping sensor only.
You also need to configure the notification trigger(s) in a way that it only triggers a template once (when the sensor state is down for x seconds). Ex:
When sensor state is Down for at least 60 seconds, perform NotificationTemplate When sensor state is Down for at least 300 seconds, perform No notification and repeat every 0 minutes When sensor state is no longer Down, perform No notification
If you don't want to use the inheritance for the notification trigger, you could create a Library and add the devices there. This would allow you to configure a notification trigger only for them.
Please, let me know if I'm on the wrong way here.
Regards.
May, 2022 - Permalink