This discussion has been locked. The information referenced herein may be inaccurate due to age, software updates, or external references.

You can no longer post new replies to this discussion. If you have a similar question you can start a new discussion in this forum.

Is there a way to set a pre-condition on a subset of interfaces?

JaredC over 13 years ago

Hi everyone, I've been using Orion now for a little over a month now and I'm still learning all the tweaks and settings that are possible.

I'm hoping for some insight on how best to implement this, as I'm sure it's possible in many ways.

In one of the several data centers that Orion is monitoring, local IT staff have access to the physical interfaces on the Access layer Cisco switch. This specific data center has seen a lot of activity lately, and I'm having to go in and "check" the switch for new ports -- ports without descriptions that are showing connected.

I'd like to be able to have Orion step in and assist. What I'm thinking of is to set all of the unused ports to "unplugged" instead of ignoring them.

Then, I'd like to receive an email when the port goes from "unplugged" to "up"

Sounds easy, but I cannot see to find the missing piece to only look at unplugged ports. My first attempt netted me over 7200 alerts in my inbox as the reset of "up" was triggered by EVERY port on every node Orion is monitoring.

My current plan is to create a basic alert. Under Monitored Network Objects, select only the unplugged interfaces. The alert trigger will be up and the reset will down. But this still means I'm going to have to manually change the alerting structure each time an unplugged interface is used.

In summary, I am attempting to do the following scenarios simultaneously:

Scenario 1: Node is up -> Interface is up -> Interface goes down -> email is sent. Interface comes back up -> reset email is sent.

Scenario 2: Node is up -> Interface is unplugged -> Interface changes to up -> email is sent.

Seemless integration of interfaces that are monitored in Scenario 1 that were activated via Scenario 2.

So, what do you think?

Thanks,

-JaredC

0 mcbridea over 13 years ago

I would recommend using advanced alerts rather than basic ones. Scenario 1 is just an interface down alert. This is one of the out-of-the-box advanced alerts. For scenario 2 it think this will do the trick, but I have not tested it. You will have to make sure the datacenter is either in the config as the location or create a node custom property and assign it to the data center switches. Then add the node custom property into the alert where I have placed the location qualifer.
Cancel
Vote Up 0 Vote Down

Cancel
0 JaredC over 13 years ago in reply to mcbridea

Hi, and thank you for the suggestions. I have tried several variations over the last few days and have not yet achieved what I'm hoping to accomplish. Thinking it's just something simple, like I'm using the inverse instead of the converse.
The unplugged interfaces need to send an alert only when they go up for the first time. Once that occurs, I will go in and manually reconfigure the interface to no longer consider down as unplugged. So if the trigger is "up" with an email for such a trigger, the reset becomes unnecessary. I attempted to make the trigger and the reset the same, but that supressed the email action that's setup on the trigger.
I took your suggestion and created an SNMP location on the switch(es) I'm concerned with. I've made the trigger alert when all (both) of the conditions apply: "Interface status is equal to up" and "location is equal to PeanutButter" The Trigger action is to email me.
The reset condition is also "Interface status is equal to up" No additional Reset action.
Is there not a way to trigger on a pre-existing condition/variable?
Thanks again for the help...

-JaredC
Cancel
Vote Up 0 Vote Down

Cancel
0 qle over 13 years ago in reply to JaredC

No, I don't believe there's a method to trigger on a pre-existing condition. However, to address your question about the alert, I don't think you should set the reset condition to match the trigger condition. As you've noticed, your trigger actions won't fire. Just set the reset condition to anything else like interface status is down. It really doesn't matter since we're primarily interested in the trigger condition.
Now back to your trigger conditions: you'll notice that Andy (aka mcbridea) suggested three trigger conditions but you only mentioned two of them. Remember that the third condition ("Interface Caption is empty") is important since you're looking for "ports without descriptions".
Once the trigger conditions are setup properly, you'll get an e-mail of all interfaces that have a blank/empty name within NPM if they're currently up. It doesn't matter if the alert is reset since you'll have an e-mail record of those interfaces.
Hopefully, this makes sense and helps you.
Cancel
Vote Up 0 Vote Down

Cancel