[Nagiosplug-devel] check_ntp plugin on the same host.

[Subhendu Ghosh]

On Tue, 10 Feb 2004, JAMES BOYLE wrote:

> Hi,
> 
> I was thinking that adding an option to the check_ntp command to not
> count an offset of 0.000000 when ntpd is local, as an error.  To explain
> what I mean a little better...
> 
> System Alpha:  has NTP, stratum 2, and Nagios
> System Bravo: has NTP, stratum 2
> System Charlie: has Nagios
> 
> The Nagios on Alpha will report an error with check_ntp -H <Alpha> when
> the offset is exactly zero, while the Nagios on Charlie reports there is
> no problem with check_ntp -H <Alpha>.
> 
> --James
> 

I had posted another question aabout a month back that is somewhat 
related...

check_ntp uses 2 methods: "ntpdate" and "ntpq"

For systems running a full NTP server, "ntpq" gives us more accurate 
information like stratum sync and jitter.  Here I don't think we should 
run ntpdate as well.

On systems running SNTP, the ntpq command does not work, but "ntpdate" can 
be used to get an offset relative to the localhost.

I wanted to add a switch to use either ntpq (for NTP server) or ntpdate 
(for SNTP servers)

With the switch and using ntpq only, the offset is not an issue...

I didn't get any feedback on that - I think I'll go ahead and patch it
with default behavior going to ntpq.  This will break existing usage 
against any Windows boxes....



-- 
-sg