David Lord wrote:
Mark C. Stephens wrote:
This is what I had with no time1:
remote refid st t when poll reach delay
offset jitter
==============================================================================
oPPS(0) .PPS. 0 l 13 16 377 0.000
-0.007 0.005
*GPS_PALISADE(0) .GPS. 0 l 12 16 377 0.000
0.008 0.003
And this is what I get with time1 of 0.015
remote refid st t when poll reach delay
offset jitter
==============================================================================
oPPS(0) .PPS. 0 l 10 16 377 0.000
0.012 0.005
*GPS_PALISADE(0) .GPS. 0 l 10 16 377 0.000
15.031 0.008
Seems I have put myself out 15 ms by adding time1 0.015
Also in my ntp.conf I have:
saveconfigdir /etc/ntpd
however if I do a ntpq -c saveconfig it comes back asking for a filename.
Have I not compiled saveconfig mechanism in or something?
.....
You don't need to get a very accurate time from the GPS as the
PPS when present sets the offset. Some GPS are a disaster when
used to set a clock and several of my devices have variations
of > +/- 50 ms. Your GPS_PALISADE is very accurate by comparison.
eg.
fudge 127.127.28.0 stratum 4 time1 0.024 refid MSFa
fudge 127.127.20.2 stratum 6 time2 0.417 refid GPSb
Note that I don't really want either to be used for offset as
neither is as good as internet time.
If you really want to fudge out the time offset of GPS_PALISADE
you would need 0.000008 (s) or better run peer_summary for a
while so you can get an average in case that value varies.
Hi
from my peer_summary
ident mean(ms) rms max(ms)
127.127.22.2 (.PPSb.) 0.000 0.004 0.004
127.127.20.2 (.GPSb.) -0.125 20.180 69.578
Peer and loop summary are scripts from the ntp distributions.
4.2.6p5 has both a perl and .sh versions.
David
_______________________________________________
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions