Nagios Support Forum

Guys/Gals,

I have a cross system topic. I'm trying to create system alerts and ship them to our nagiosXI server via NRDP.
So far I can get the alerts to show critcal locally when I manually run them. I have a proper host, token and service. I even tested to make sure they were getting there by changing the hostname to something different. It then subsequently showed up in the NagiosXI server as an unconfigured object, as expected.

Any ideas on where to start troubleshooting this?

-
Wayne

I guess I'm not sure what the problem is, since you say the Unconfigured Objects part is expected. Can you elaborate?

Wow. I can't believe I forgot to include the problem. Just chalk that one to me having a "moment"
The issue is that NagiosXI doesn't reflect the error condition or send an alert based on the condition.

--
Wayne

What versions of Log Server/XI are you running? I've tested this in the past with good success.

Nagios Log Server - 2015R2.2
NagiosXI - 2014R2.7

I just tested this on a Log Server 2.2 paired with an XI 2.7 installation - critical alerts come through as critical events appropriately. Can you show me how you've defined the alert in your Nagios Log Server 'Alert' Dashboard? I'm not sure what the problem might be here.

This is what I'm seeing on my end.
I did see something you had done differently in your setup. I was trying to use an existing service to alert on. However even when I created a new passive check with a separate name for the test alert. The status still doesn't make it across.

Here are some screenshots of the configuration on both ends.

If you click the 'run alert now' button, does the status of your 'Nagios404Test' check change at all? If not, the only difference between our setups is that I'm using http as opposed to https. Can you access the https nrdp interface and use your token properly?

Using a web browser, navigate to:
https://nagiosxiserver/nrdp

Try entering your token and verify that the token takes successfully. Is it possible for you to try http as a test measure? I don't know why else this wouldn't be working.

Thanks that got me to the bottom of it.

I went to go check my NRDS config manager. Which was now strangely missing. It appears the problem is when you edit the configuration in the NRDS config manager and then add a token through the inbound interface, it messes up the /usr/local/nrdp/server/config.inc.php token section.

It appears to terminate the string for one token while leaving the unknown token in the opposing configuration line.

This is in the "$cfg['authorized_tokens']=array(" section