Dependency aware Alert examples

Hi guys,

Could anyone please share screenshots of their trigger condition for an alert that is suited to mute Down alerts on Child objects that are down?

I have a dependency where a router is the parent and a group is the child. I only wish to receive alerts when the router is down and suppress the hundreds of alerts I would get when the group members are down. Any help greatly appreciated.

Top Replies

  • This doesn't impact the alert definition as the related entity never goes into a 'Down' state when its parent is down. Instead, the entity goes into an 'Unreachable' status. So unless you've configured your alerts to trigger when the status is anything other than 'up', you won't be alerted when an entity is 'unreachable' (parent is down) rather than 'down' (parent is up, child is down).

  • Hi aLTeReGo

    Thanks for the swift reply. There's something I am missing here for sure, so I hope you can explain it further for me.

    This is what I have done to test the functionality:

    I've created a static group that includes some nodes that are Up and confirmed that it works as it should.

    I then create a manual dependency where I selected a random node that was down, set it to be the Parent and selected the static group that includes Up nodes to be the Child.

    At this point, I was expecting the Up nodes from the group to have an Unreachable status but that wasn't the case. Reviewing the nodes in Summary views and even their respective Details view still showed them as being Up.

    Nowhere was I able to see the Unreachable status.

    Was my test procedure configured wrong or could there be an issue with my installation?

    Also, does having automatic dependencies enabled break this test, and if so, does that mean that Automatic Dependencies must not be enabled if you want something like this to work?

  • 'Unreachable' is a substitute status only for 'down'. If the parent is 'Down' but the children are still up and polling correctly, their true/real status will be reflected.

  • Does that mean that I'll have to use a specific condition in my Node Down type alerts or will the standard Node status is equal to Down condition work here?

  • Hey Deltona what it means is that your test scenario is slightly wrong, this is a mockup from my lab:

    pastedImage_17.png

    So my Router is the Parent, and my Group is the Child. If my Router drops I will lose connection to my eMail System Nodes and I won't get Node Down Alerts for those devices.

    The caveat here as aLTeReGo pointed out if your Alert condition is NOT Node Status = Down, then this won't work.

  • I am confused. AFAIK my test scenario as described resembles that in your lab.

    Your parent object, the router, is my down node. Your child object, the eMail System group is my group of up nodes.

    So why is it not working? Do you have a csreenshot of the Trigger Condition that would work for this?

  • Hi Deltona,

    I admit this is hard to simulate because I can't force my WAN Router down lol. I'll see if I can mock up a different one later on. Let me jump over to my Lab system and see if I can get a screenshot.

  • You might be able to simulate wan router down.

    In windows firewall on the orion server, add a rule to block all traffic to that IP.

  • Hi David,

    Does it actually have to be a wan router? Will it not work with any other node that is or is set to Down state manually like when the IP is blocked locally in the Windows firewall?

  • It doesn't have to be a WAN Router, but it's hard to simulate because the end devices need to be behind the Parent device. If SolarWinds can still ping the device directly then it will register the true status (as per your existing test).

    This is my alert trigger (ignore the Scope part as it's a Demo lab)

    pastedImage_0.png