Skip to main content

 

Hello everyone,

I am creating a new topic because, despite multiple readings of the documentation on notifications, I still can't refine my notification settings as I would like.

Currently, my hosts (e.g., a server) need to send me an alert immediately when they stop responding. I have configured the notifications like this, and it seems to be working quite well.

 

Our biggest problem now is the service notifications. Indeed, as soon as a server's memory is saturated or the CPU is saturated for 15 minutes, I receive an alert. With the number of servers we have, this quickly becomes unmanageable.

I know on which servers it is “normal” to reach these thresholds, and I would like to receive a single alert, for example, only if the 95% threshold is reached for more than an hour.

Similarly, I have a disk on a server that is always at 98% capacity. I would like to stop receiving a daily alert about this.

Here is how I have configured the notifications for RAM memory, for example:

 

For example, here, I receive an alert almost every day, whereas I would like to receive the alert only once when the critical threshold is reached and then again when the server's memory returns to “OK” status.



Thank you in advance for your help.

I think what you are trying to do, would be configured in the Escalation-Setting:

 

 

If you configure “First Notification” as “1” and “last Notification” as “1” you will only get ONE Message regarding the Problem. Also click on the Question-marks - they have quite the handy descriptions in there.

 

Regards

Timo

p.s.: For different behaviour -> services in different groups and more than one escalation with specifica per Service-Group ;)

Hi @regis_shr you can use “ First notification delay”, if the resource is still in a non-ok status after this delay, you will receive your notification. If the resource go back to OK during this delay, no notification will be sent.

Thanks @TA-FFM  for the feedback,
I have configured the escalation as follows:

 




Should I enable the inheritance option?
With this escalation configuration, will I receive only one notification for the hosts and services?

Thanks! :)

Hey, 

 

first one and last one set to “1” means you only get that one notofocations, yes.

 

You can do other things with that feature too. 

For example: One group is getting firts=1 and last=0, intervall=5 - this group gets Notifications beginning with the first and will not stop, getting a new one everay 5 minutes until the Alarm is acked.

 

Another team can now have settings like: first=10, last=0, intervall=5 which means if no one from group one acknowledges the alert the 11th Notification will now also be sent to the secondf team together with the first team until it is acked.

 

first, last etc. always means the first, second, third iteration of that alarm.

 

Also see what Laurent wrote about the delay if you donÄt want to be that “trigger-happy”.

 

The most important thing. You can have hosts/services in different grouping and use that to put them in differently configured escalation-settings.

Sry for the inheritance, don’t use it much. Maybe Laurent can elaborate on that for you.

 

Regards

TAFFM

Thank you to @Laurent  and @TA-FFM for your help. I find the notification system quite complex to understand, but I imagine it is designed to optimize the management of the alerts we receive.

Reply


Terms & ConditionsCookie settings