[ntp:questions] New to NTP on Linux
Brad Knowles
brad at stop.mail-abuse.org
Tue Nov 2 12:17:49 UTC 2004
At 9:56 AM +0000 2004-11-02, Dusty Bin wrote:
> I realise that the Garmin GPS25LVS isn't ideal, but it would seem
> suprising that the 1PPS signal is actually as inaccurate as indicated
> by my current results.
I had vaguely recalled seeing some pages which said that the
Garmin units were pretty bad at handling NMEA/PPS, but could not
remember where I had read that. With a bit of Googling, I found the
page at <http://www.jensar.us/~bob/garmin/> which appears to agree
with my memory of accuracy problems with these units, but this isn't
the page I recall having seen before.
Although they are more related to positional accuracy than timing
accuracy, I also turned up pages at
<http://www.edu-observatory.org/gps/gps_accuracy.html>, and
<http://www.wsrcc.com/wolfgang/gps/accuracy.html>.
I found other pages which put PPS accuracy of Garmin units in the
one microsecond range, whereas we know from recent discussions on
this list that the Oncore devices are in the 50-80 nanosecond range.
I did some of my own tests with the public list of stratum one
timeservers which are supposedly known to be running Garmin or Oncore
devices, and didn't see any significant performance differences
(information gathered with "ntpq -c rv" and "ntpdc -c sysinfo"
against the named server, with the jitter and stability numbers
coming from ntpdc).
I compared the jitter and stability numbers returned by "ntpq -c
rv" against those returned by "ntpdc -c sysinfo", and they seemed to
be significantly different, even for the same server. I wasn't able
to confirm which was more accurate, so I arbitrarily decided to use
the jitter/stability data from ntpdc.
Anyway, here's what I found on my survey:
Server: ntps1.pads.ufrj.br
System peer: GPS_NMEA(0)
Reference ID: [GPS]
Device type: garmin
Version: ntpd 4.2.0 at 1.1161-r Tue Oct 21 01:35:24 BRST 2003 (1)
Operating System: Linux/2.4.27-NANO
Processor: i686
Jitter: 0.000000 s
Stability: 0.002 ppm
Server: ntp.cesnet.cz
System peer: GPS_NMEA(0)
Reference ID: [GPS]
Device type: garmin
Version: ntpd 4.0.99k23 Tue Jan 15 11:15:28 CET 2002 (7)
Operating System: Linux2.2.19-NANO
Processor: i586
Jitter: 0.000000 s
Stability: 0.005 ppm
Server: tick.mhpcc.hpc.mil
System peer: GPS_VME(0)
Reference ID: [USNO]
Device type: oncore
Version: ntpd 4.2.0 at 1.1161-r Wed Apr 28 22:57:21 UTC 2004 (19)
Operating System: HP-UX/B.11.11
Processor: 9000/800
Jitter: 0.000015 s
Stability: 0.006 ppm
Server: ntp0.mcs.anl.gov
System peer: GPS_VME(0)
Reference ID: [USNO]
Device type: oncore
Version: ntpd 4.2.0 at 1.1161-r Thu Jul 29 23:29:18 UTC 2004 (20)
Operating System: HP-UX/B.11.11
Processor: 9000/800
Jitter: 0.000000 s
Stability: 0.000 ppm
Server: bonehed.lcs.mit.edu
System peer: GPS_PALISADE(0)
Reference ID: [CDMA]
Device type: oncore
Version: ntpd 4.1.0-a Wed Jul 7 17:09:19 EDT 2004 (1)
Operating System: FreeBSD4.9-RELEASE-p5
Processor: i386
Jitter: 0.000015 s
Stability: 0.000 ppm
Server: ntp.quidnet.com
System peer: GPS_ONCORE(0)
Reference ID: [GPS]
Device type: oncore
Version: ntpd 4.1.1b at 1.829 Thu Apr 10 23:44:47 EDT 2003 (1)
Operating System: Linux2.4.20-nano
Processor: i686
Jitter: 0.000000 s
Stability: 0.030 ppm
Server: time.keneli.org
System peer: GPS_ONCORE(0)
Reference ID: [GPS]
Device type: oncore
Version: ntpd 4.1.2a at 1:4.1.2a-2 Fri Mar 19 00:41:37 EST 2004 (1)
Operating System: Linux2.4.25-tick-1
Processor: i686
Jitter: 0.000000 s
Stability: 0.006 ppm
Server: time.xmission.com
System peer: GPS_ONCORE(0)
Reference ID: [GPS]
Device type: oncore
Version: ntpd 4.1.1a at 1.791 Fri Oct 18 16:01:30 MDT 2002 (3)
Operating System: Linux2.4.18-nano
Processor: i586
Jitter: 0.000000 s
Stability: 0.002 ppm
--
Brad Knowles, <brad at stop.mail-abuse.org>
"Those who would give up essential Liberty, to purchase a little
temporary Safety, deserve neither Liberty nor Safety."
-- Benjamin Franklin (1706-1790), reply of the Pennsylvania
Assembly to the Governor, November 11, 1755
SAGE member since 1995. See <http://www.sage.org/> for more info.
More information about the questions
mailing list