Alerts based on UnDP

Can you share the following?

- Your NPM Version

- Your UnDP's name.

- Your network equipment vendor that you're polling against. (Cisco? Juniper?)

- Your UnDP's polled OID.

With this, the community (or myself) could potentially build you a sample SQL query to find the results you need. That, and I want to steal this idea, because I've had this problem before.

Hi dhanson, feel free to steal, share and investigate.

whatever to help me and other, that's why we´re here I guess :-)

I´m using NPM 11.5 and polling Cisco equipment.

For stratum value I´m using OID 1.3.6.1.4.1.9.9.168.1.1.2 (cntpSysStratum) ...assuming this is the UnDP name you asked for

This returns a value.

For NTP state I was going to use OID 1.3.6.1.4.1.9.9.168.1.1.11 (cntpSysSrvStatus)

...but never got the boxes to return a value.

I have been able to set the critical and warning thresholds of the UnDP.

And I thought when creating an alert you could reference a UnDP, but does it require some tweaking under the hood?

Thanks

Ok, couple things.

1. Based on your statement, I have a potentially easy question for you: Did you make these with the Universal Device Poller application on your SolarWinds server? The rest of my questions are under the assumptions that this is true.

2. Double check the advanced options on your UnDP. Mine are:

MIB Value Type: Raw Value

Format: Enumeration

SNMP Get Type: GET NEXT (I use SNMPv3)

Polling Type: Node

Polling Interval: <blank...it defaults to standard poll rate>

3. Also check and see if you are keeping historical data. I think you are already, because you said you had results in your graph, but worth double checking.

4. For your alert trigger condition, check that the type of property to monitor is set to "Custom Node Poller". By default, this is typically set to "Node".

5. For the trigger, use syntax like this (I still use 11.0.1, bit scared of 11.5.2, so the syntax might be different for you):

Trigger Alert when ALL of the following apply:

     Poller Name is equal to <your UnDP's actual name>

     Numeric Value is greater than or equal to <minimum acceptable stratum, typically 3>

If this doesn't work, let me know.

Now for the NTP State OID:

None of my devices support this OID. I'd walk your devices to verify they return a value through this OID, and if they do, follow the same process, but you might want to tweak the advanced settings on the UnDP a little bit, or for the alert, have it reference the text string instead of Numeric Value.

-HTH!

oskare, will your alerts have actions?

Would you be happy if you had a page showing UDP values outside of certain thresholds?  That can be done without alerts.

Yes, that might be useful.

...at least worth a look

Sent from Samsung Mobile

dhanson‌

Some success, my first advance alert config and missed the option "use value/macro"

So now I am getting alerted on the UnDP,  But always alerted now, decpite if value is 3 or 16 :-)

But the two nodes with this poller assigned are both alerting, one has 16 the other has 3.

However one is showing red leaf icon the other a green leaf, plus a few ciscoEnvMon where triggered on other nodes?!?!

I will look at this tomorrow when I have more time. Too tired now.

-reg.

Oskar

"now I am getting alerted on the UnDP, but always alerted now" - This might have to do with what you're triggering on. Can you send some SS's of your configured alert and action?

ciscoEnvMon alerts are based on either temperature, power, or internal hardware. If you're getting these alerts, I'd definitely investigate those as a separate issue.

Precisely what I needed to get this working. Attached is a screen shot of how I set the trigger up for anyone who needs an assist. The trigger I set up was to alert on a Universal Device Poller for a TrippLite environmental probe for humidity out of bounds (>90%).