Different notification for HOW service fails?

[dunnlow]

We have issues concerning monitoring linux services on different hosts. We notify certain people when a service fails. The problem is that the notification looks the same (goes to the same person) regardless of whether the failure happened because the nagios box couldn't reach the host to check or when the service connected to the host successfully for the test actually ran and failed. This is important to use because if the service fails due to a connection issue, we want to notify a different person than if the specific test ran and failed. Is there a good way to handle this?

Thanks!

[slansing]

Ideally your host check for that server will go out more often than the service checks, you can assign other contacts to the host check and they will be notified if the server is unreachable. Your service contacts will then be notified if the check itself failed on the service level. Also, keep in mind that if the service check fails due to the host being unreachable the error code will change to something akin to a timeout error message.