[ntp:questions] Re: Please demistify the SHM driver

Paul Croome Paul.Croome at softwareag.com
Thu Sep 25 08:30:05 UTC 2003

Jerome Oufella <oufella at SPAMBLOCKchez.com> wrote in message news:<bkthjo$b45$1 at news.univ-mrs.fr>...

> Paul,
> The LOCAL driver is only used as fallback, if the serial GPS (rare but 
> possible) or network link ever fall down (this often happens).
> My GPS clock sends an RS232 frame once every minute. The frame 
> informations give date, hour, minute and seconds only (not msec, nor 
> usec). I just hope its internal usec and msec are constant when the 
> RS232 frame is emitted. Maybe not, this may explain the jitter.
> The previous 4.625 value was acquired after only some dozens of minutes 
> of work.
> The actual values (after some days up) are :
> remote      refid    st t when poll reach delay   offset jitter
> ==============================================================================
>   LOCAL(1)   LOCAL(1) 9  l 17   64   377   0.000   0.000  0.008
> *SHM(1)     SHM(1)   10 l 11   64   377   0.000   1.349  0.044
> This seems better.
> Thanks again.
> Jerome.


If the GPS becomes unavailable or if the NTP server becomes unavailable
because of LAN problems, NTP will continue to free-run, using its current
estimates of clock offset freq etc. You don't have to configure the local
clock driver. Also, I find it strange that you have the GPS refclock
at stratum 10 and the local clock at stratum 9.

As others have written, for best precision, a PPS interrupt signal is required.
However, if you don't have that...


More information about the questions mailing list