[ntp:questions] NTP and GPS

Miguel Gonçalves miguel.barbosa.goncalves at gmail.com
Fri May 3 10:47:39 UTC 2013


Hi!

Many thanks for your suggestions!

In the meantime I've just bought a Trimble Acutime Gold and used a
RS422-to-RS232 converter and apparently it is working fine on a FreeBSD 8.3
machine (NTP 4.2.7p367).

I am getting a bit of jitter on the serial port. On FreeBSD 7 there was the
possibility of disabling the UART's FIFO but I can't seem to find how to do
it on FreeBSD 8.

Any help would be greatly appreciated.

Here is some more relevant information about the machine and set up:

--- /etc/ntp.conf ---
server a.b.c.d iburst
server k.l.m.n iburst

server 127.127.29.0 mode 3
fudge 127.127.29.0 minpoll 4 maxpoll 4
fudge 127.127.29.0 flag2 0

statsdir /var/log/ntp/
filegen peerstats file peers type day link enable
filegen loopstats file loop type day link enable
filegen clockstats file clock type day link enable
--- /etc/ntp.conf ---

acutime# ntpq -p
     remote           refid      st t when poll reach   delay   offset
 jitter
==============================================================================
*GPS_PALISADE(0) .GPS.            0 l    2   16  377    0.000   -0.248
0.034
+a.b.c.d         w.x.y.z          2 u   51   64  377   10.941   -0.520
0.425
+k.l.m.n         w.x.y.z          2 u   51   64  377    9.201   -1.984
0.410

--- /var/log/ntp/loop ---
56415 38463.327 -0.000358705 144.772 0.000047732 0.005580 4
56415 38479.326 -0.000290738 144.768 0.000050705 0.005449 4
56415 38495.326 -0.000327819 144.763 0.000049209 0.005396 4
56415 38511.327 -0.000327237 144.758 0.000046050 0.005347 4
56415 38527.327 -0.000257539 144.754 0.000049626 0.005191 4
56415 38543.327 -0.000308850 144.749 0.000049840 0.005134 4
56415 38559.326 -0.000247656 144.745 0.000051397 0.004984 4
56415 38575.327 -0.000247452 144.742 0.000048096 0.004850 4
56415 38591.327 -0.000281144 144.737 0.000046540 0.004784 4
56415 38607.327 -0.000318896 144.732 0.000045534 0.004794 4
56415 38623.327 -0.000256571 144.729 0.000047956 0.004694 4
56415 38639.326 -0.000294313 144.724 0.000046801 0.004669 4
56415 38655.327 -0.000294382 144.720 0.000043799 0.004646 4
56415 38671.327 -0.000228657 144.716 0.000047101 0.004518 4
56415 38687.327 -0.000268336 144.712 0.000046239 0.004467 4
56415 38703.327 -0.000301464 144.707 0.000044810 0.004485 4
56415 38719.326 -0.000236024 144.704 0.000047878 0.004384 4
56415 38735.327 -0.000244381 144.700 0.000044883 0.004307 4
56415 38751.327 -0.000279280 144.696 0.000043759 0.004301 4
56415 38767.327 -0.000237060 144.692 0.000043570 0.004221 4
--- /var/log/ntp/loop ---

What can I do to improve this?

Thanks in advance for any help!

Kind regards,
Miguel

On Wed, May 1, 2013 at 8:11 AM, David Woolley <
david at ex.djwhome.demon.invalid> wrote:

> Miguel Gonçalves wrote:
>
>
>> If one of the stratum 1 goes nuts the Cisco will mark it as a falseticker
>> and discard it's data, right? I believe the Cisco will then choose the
>> other stratum 1 to sync from. Right?
>>
>
> Noting the lack of sufficient votes for a falseticker, already mentioned.
>  ntpd will normally synchronise to both stratum one servers, and, with a
> lower weighting, to any higher stratum servers, although it will only
> report the error statistics and source identity of one of them.
>
>
>> If the Internet connection is overloaded (not usual here) the Cisco will
>> choose predominately the stratum 1 servers. Right?
>>
>> Is this circular topology (a client using a server that in turn references
>> the other two sources the client is using) a good idea?
>>
>> I know a better approach would be to add another stratum 1 (having 3 in
>> total) and ignore completely the Internet servers. Can I really get away
>> with the above configuration? I am not looking for the ultimate precision
>> but I intend to use these two receivers I have to their fullest.
>>
>> How good is NTP detecting a problem with a reference clock? Let's imagine
>> there's a problem with one particular satellite and makes the time jump a
>>
>
> GPS time solutions are normally made from multiple satellites (although in
> pre-surveyed mode, a time receiver could use just one.
>
>  few mili-seconds. With only one reference clock will NTP follow blindly or
>> ignore it? Adding another GPS would be pointless but a radio receiver
>> would
>>
>
> Stratum one servers normally set a root dipersion that would mean that a 1
> millisecond error would still be within the allowable error bounds, even
> without the increases to those error bounds to allow for time since last
> measurement and network propagation delays, so it is very unlikely that a
> clock that differed by only or two milliseconds.
>
> In any case, if the radio clock reports a fault, the stratum one server
> will continue to serve time, but with an increasing root dispersion.  It
> will get ignored when that root dispersion gets too high.
>
> ______________________________**_________________
> questions mailing list
> questions at lists.ntp.org
> http://lists.ntp.org/listinfo/**questions<http://lists.ntp.org/listinfo/questions>
>


More information about the questions mailing list