check_ntp

[Demetri Mouratis]

On Tue, 7 Sep 2004, Andreas Ericsson wrote:

> Leif Nixon wrote:
> > "C. Bensend" <benny at bennyvision.com> writes:
> >
> >
> >>>The plugin considers an zero offset to be erroneous. I don't know why.
> >>
> >>Check the archives; there was a lively discussion about exactly this a
> >>month or two ago.
> >
> >
> > I know. We never did get that detail sorted.
> >
>
> check_ntp thinks an offset of 0 is erroneous because many ntp servers by
> default fetches time from the local system if it can't connect to the
> preferred NTP server. A 'true' offset of 0 (where time is actually
> compared to that of its higher stratum server) is very, very rare. It is
> more likely the result of imprecise system calls than the real status.
>
> Nevertheless, it does happen (for one reason or another), so it would be
> nice to add an argument to skip it complaining about this. I'll get
> going on that anyday now. ;-)
>

A zero offset can by erroneous, but is not necessarily erroneous.  Look at
the stratus number in conjunction with the offset and you can make a more
accurate determination.  Zero offset and stratus 16 is probably
erroroneous.

We use 1.4.0alpha1 from Dag's nagios-plugins-1.4.0alpha1-1.  We get zero
offset all the time in house.  We have a two-tier NTP configuration with
Nagios checking all the Linux hosts on the network pulling time from a
Cisco box pulling from a tier 2.

On the client networks in the field, the two tier-NTP returns but most
their NTP servers turn out to be Windows boxes.

I have addressed the problem in house by increasing max_attempts for ntp
service checks to 5.  While we get intermittent zero offset, the five
checks lets time drift few milliseconds which puts the check back in OK
state.

In the field, we get continual CRITICALs due to high offset.  The
customer's NTP servers are not doing their job.
---------------------------------------------------------------------
Demetri Mouratis
dmourati at linfactory.com



-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
Nagios-users mailing list
Nagios-users at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nagios-users
::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. 
::: Messages without supporting info will risk being sent to /dev/null