Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 13 Mar 2014 11:01:14 -0700
From:      Charles Swiger <cswiger@mac.com>
To:        mexas@bris.ac.uk
Cc:        FreeBSD - <freebsd-questions@freebsd.org>
Subject:   Re: ntp frequent time resets - battery dead?
Message-ID:  <F146857D-4705-4DDE-8C85-F712D311BEF0@mac.com>
In-Reply-To: <201403121926.s2CJQXbQ078112@mech-cluster241.men.bris.ac.uk>
References:  <201403121926.s2CJQXbQ078112@mech-cluster241.men.bris.ac.uk>

next in thread | previous in thread | raw e-mail | index | archive | help
Hi--

On Mar 12, 2014, at 12:26 PM, Anton Shterenlikht <mexas@bris.ac.uk> =
wrote:
>> Yes, it looks to be about ~1 second per hour.  That's 1:1000 ratio,
>> which is getting close to the typical kernel limit on adjtime().
>>=20
>> Tweaking the step threshold might help.  Or look into tickadj / =
ntptime.
>=20
> Sorry, I don't understand this.
> Please elaborate.

OK.  ntpd has two ways of adjusting the clock; either it gradually slews =
the time
via adjtime() or it jumps the clock via settimeofday() and logs the =
"time reset"
lines you've noted.

Which method is chosen is based upon the step threshold, which defaults =
to 128 ms
(aka .128 seconds).  Increasing the step threshold will have ntpd try to =
slew
the clock even for the larger corrections your case seems to need.

Regards,
--=20
-Chuck




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?F146857D-4705-4DDE-8C85-F712D311BEF0>