No status for the error -Return code of 255 is out of bounds

[MSPk]

Hi Team,

We have configured some linux/aix servers for monitoring and getting the status message as - Return code of 255 is out of bounds. we have done some quick checks and found that the ports on the monitored servers are not open. now, our ques is the status of each of these services is blank (not warning/critical or unknown) and when I go to the service summary dashlet its tough to identify that some of these services are not working as dont come under warning/critical/unknown and are missed out from the problems column. Please suggest if there is a workaround for this or if we have done anything incorrectly.

[abrist]

Are these run through nrpe, check_by_ssh, or some type of other agent?

[MSPk]

they are run through NRPE

[abrist]

Well, port 5666 has to be open for the nrpe checks to work . Have you checked the iptables rules and made sure the nrpe/xinetd service is running?

[MSPk]

Yes, the port was open and server was pingable, however the services were showing the status message as 'Return code of 255 is out of bounds' and the status is blank shown in the figure.

Upon some investigation we found that when the nrpe service is stopped or not responding we are getting this status. What bothers is that when this error is comes up, the service ends up without any status (none of OK/Warning/Critical/Unknown) and no email is sent out for the service being down/unknown nor is it shown in the stats of service status summary. Please help.

[abrist]

You should actually get a "connection refused" it the service is stopped and the port not active. Out of bounds errors are usually due to a plugin not exiting with an exit code. Can you run the check from the cli and then check the exit code?

Code: Select all

./check_nrpe -H <host> -c <command>
echo $?

[lmiltchev]

When you run the check from the CLI, and get "255 is out of bounds" message, what is the exit status code?

Code: Select all

echo $?

[litsupport.box]

If you get this error on only one or two hosts you could remove them from nagios / remove services -> remove host / and then add it back and see what happens. I got this message when checking with vmware_api plugin with a mod gearman worker. When testing the check it works, but for some reason it gave me false/positives.. anyway you could try.

[lmiltchev]

MSPk, any updates?