[SOLVED] redirecting nagios to check 3 times each error and then trigger mail

call_krushna · 03-20-2012, 11:42 AM

Hi Team,

Need to configure my existing nagios server to trigger mail only after checking 3 times any service for warning and critical .Please let me know, what should I do to get it done .

Thanks in advance .

TB0ne · 03-20-2012, 12:33 PM

Quote:

Originally Posted by call_krushna

Hi Team,
Need to configure my existing nagios server to trigger mail only after checking 3 times any service for warning and critical .Please let me know, what should I do to get it done .

Thanks in advance .

Ok...then as you were told in your other threads, all you need to do is write the script to do what you want. You were given a skeleton plugin script to monitor Tomcat. If you want it to check three times, then put a loop in your script to check three times....doesn't get more simple than that.

call_krushna · 03-20-2012, 12:45 PM

Quote:

Originally Posted by TB0ne

Ok...then as you were told in your other threads, all you need to do is write the script to do what you want. You were given a skeleton plugin script to monitor Tomcat. If you want it to check three times, then put a loop in your script to check three times....doesn't get more simple than that.

Thanks for quicky reply .

Monitoring tomcat we can do it via script .I dont think ,I need script to change the time of triggering alert of nagios script .It is not possible to write script for every service .Please suggest ..

TB0ne · 03-20-2012, 01:15 PM

Quote:

Originally Posted by call_krushna

Thanks for quicky reply .
Monitoring tomcat we can do it via script. I dont think, I need script to change the time of triggering alert of nagios script. It is not possible to write script for every service .Please suggest ..

Yes, it's VERY possible to script a check for all your services. And if you check the Nagios documentation, you'd see a possible solution to what you're asking:

http://nagios.sourceforge.net/docs/3_0/flapping.html

arashi256 · 03-20-2012, 06:35 PM

Quote:

Originally Posted by TB0ne

Ok...then as you were told in your other threads, all you need to do is write the script to do what you want. You were given a skeleton plugin script to monitor Tomcat. If you want it to check three times, then put a loop in your script to check three times....doesn't get more simple than that.

Would you want to do that, though? Because Nagios dictates the frequency of when the script is run via the NRPE daemon. So you'd be running three return-blocking checks within the loop before the script returned the result. Unless the state change of the service you're monitoring changes *very* quickly, it won't make much difference to the result that the plugin returns by looping like that. Otherwise you'd have to put a delay in the loop that would probably cause Nagios to not get it's result back at all within the allocated time frame to be useful to it's monitoring. I would have thought it's either the flapping functionality that Nagios provides or longer/shorter scheduled calls to NRPE if the output is not guaranteed. Although, as you suggest, mail need only be sent if some threshold is reached. You could store each check per NRPE call to a temporary file and then use scripting to send a mail regardless of the current host status that Nagios displays if two of the three checks is bad (for example). That's how I'd do it, anyway.

chrism01 · 03-20-2012, 08:13 PM

I'm confused; I've got 253 service checks atm and under the 'Attempts' col, they all say 1/3 or 1/4.
I'm pretty sure that's the default for most checks, unless you wrote your own.
Nagios doesn't email me until its certain its got an issue; see templates.cfg => 'define service" for 'generic-service' => 'max_check_attempts 3 ; Re-check the service up to 3 times in order to determine its final (hard) state'
See also http://nagios.sourceforge.net/docs/n...finitions.html

HTH

polinux · 03-28-2012, 05:21 AM

You could also set up a specific notification scheme for every single service. So for example:

Code:

define service{
        use                             nrpe-service
        host_name                       machine_name
        normal_check_interval           5
        service_description             Load
        check_command                   check_nrpe!check_load
        notifications_enabled           1
        max_check_attempts              3 
        }

In this case it woild send an email after 3 checks.