Tony Hoyle wrote:
FYI I setup an old laptop with Freebsd and the jitter is about 1% of
what it was on the Linux machine (and the laptop has a much slower
I'll quantify that...
ntpq> lpeers
remote refid st t when poll reach delay offset
jitter
==============================================================================
+ntp0.ja.net .MSF. 1 u 36 64 377 17.434 1.290
0.876
+veracity.mcc.ac 193.63.106.104 3 u 48 64 377 23.238 1.233
4.898
+utserv.mcc.ac.u 128.86.8.123 2 u 48 64 377 22.868 1.505
0.133
+scarp.mc.man.ac 128.86.8.123 2 u 49 64 377 22.736 1.134
0.182
*GPS_NMEA(0) .PPS. 0 l 8 64 377 0.000 0.012
0.012
192.168.44.255 .BCST. 16 u - 64 0 0.000 0.000
4000.00
Old machine was Debian Stable, running kernel 2.6.8 + PPS Patches on an
Athlon XP 2400+, and gave GPS jitter of between 1.000 and 6.000 usually
(with PPS on - you don't want to *know* what it looked like with PPS off).
New machine is FreeBSD 6.0 on a P4 266 w/PPS enabled kernel. The
network latency is much higher on this machine, but the GPS jitter is
virtually nonexistant (note that it's synced off the GPS hence the zero
offset).
It's be interesting to see a Linux box with a 2.4 kernel + the 'full'
PPS patches to see what that does.. I'd expect something like what the
FBSD box is doing. Don't have any old 2.4 distros around though to test
(or free boxes to test on any more).
Tony
_______________________________________________
timekeepers mailing list
[email protected]
https://fortytwo.ch/mailman/cgi-bin/listinfo/timekeepers
