[Nagiosplug-help] nrpe check of remote host runs manually, but not when scheduled by Core 3.0.3 server

[Lylex Ryan]

>From my Core 3.0.3 Server on Solaris 9, I can manually run an nrpe check of a remote host just fine:
/usr/local/nagios/libexec/check_nrpe -H <remote host> -c <remote check>

So I figure on the remote host (Centos 5.5), the nrpe daemon (v2.12) is ok, ssl checks are ok, /etc/xinetd.d/nrpe is allowing my server in, etc.

But oddly, when I let the Core Server schedule the check, the Core web interface says "Connection refused by host"

No clues in the Server's nagios.log other than the "Connection refused by host".  No clues on the remote host in the messages or secure logs, despite "debug=1" in nrpe.cfg and fresh restarts of xinetd (though all the expected info does show up in messages when I do the check manually).

I'm out of things to check.  nrpe was installed from the XI library download. Scheduled nrpe checks from the Server to other remote hosts continue to work just fine.

Thanks for any advice....Lyle