[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [gpsd-users] falsetick when only GPS is a source of time
From: |
Gary E. Miller |
Subject: |
Re: [gpsd-users] falsetick when only GPS is a source of time |
Date: |
Tue, 5 May 2015 11:07:18 -0700 |
Yo Jon!
On Tue, 5 May 2015 13:45:27 -0400
Jon Schlueter <address@hidden> wrote:
> If you are looking to do a standalone NTP server you might want to
> look at Chrony. I know a number of people have recommended it as
> better for standalone time server.
I agree with Jon, go to chronyd for standalone mode.
ntpd requires at least 3 chimers and refclcocks before it will steer
the local clock.
ntpd has an 'orphan' mode, but I would avoid it:
http://www.eecis.udel.edu/~mills/ntp/html/orphan.html
RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97701
address@hidden Tel:+1(541)382-8588
>
> Jon
>
> On Tue, May 5, 2015 at 1:24 PM, Askar Ibragimov
> <address@hidden> wrote:
>
> > Hello,
> >
> > I had interesting effect, I connected an UBLOX-6 GPS to gpsd and
> > everything worked fine at first. I only had gps, not NTP source at
> > that time. So I have typical GPS and PPS sources.
> >
> > Eventually both went 'falsetick'.
> >
> > I messed with fudge parameter and added NTP support. Then it
> > started to be ok (sys.peer state) showing jitter ~0.01 for PPS.
> >
> > I intent to make a system that is as much as possible independent
> > from internets. So I appreciate any input how to tune up the ntp
> > config such that I have no falsetick situation when I have only GPS
> > time sources. So if anybody experienced that falsetick effect and
> > know what causes it, please advice!
> >
> > Best regards, Askar
> >