[ntp:questions] Stick to PPS, even if the prefer server fails

Unruh unruh-spam at physics.ubc.ca
Thu Mar 26 02:39:26 UTC 2009


"alkopedia at googlemail.com" <alkopedia at googlemail.com> writes:

>On Mar 26, 12:48=A0am, Unruh <unruh-s... at physics.ubc.ca> wrote:
>> However if the only concern was losing PPS discipline due to prefer peer
>> problems, then I agree this is tangential.

>Yes, it was ;-)

>> >I think he's monitoring the self-reported offsets between the NTP
>> >disciplined clock and the local GPS receiver. =A0Network delay would not
>> >be a factor if so.
>>
>> I got the impression that he was using this machine to monitor a whole
>> bunch of other machine distributed over the USA to make sure that their
>> clocks were well disciplined by their own on board GPS receivers.

>You're both right:
>When I was talking about the -5 us to +5 us offset, I was talking
>about the self reported offset of the NTP disciplined clock and the

And my tangential comment was precisely about this \pm 5usec offset. I
believe that much of this is caused by the poor response of ntp to
temperature induced drift changes (but I have nothing better to suggest,
except perhaps the versions of ntp that were changed to also use the temp
to predict the drift rate and compensate automatically).
  Now you may not care in which case again my comment is tangential.


>local and the local PPS signal.
>But Unruh is right. The main task of the whole system is to monitor a
>bunch of machines (called SSU) distributed over Germany to make sure,
>that they _provide_ the right time to their clients. Of course it is
>not possible to monitor them with microsecond resolution but that's
>not needed. Just imagine one of these situations:
>- SSUs have a firmware bug, which leads to wrong timings
>- SSUs have a bug and don't notice a leap second
>- SSUs have a problem at receiving the GPS signal
>All these things may lead to wrong timings in the network and probably
>nobody will notice fast enough.

>And of course for my monitoring system it is very unlikely that GPS
>and DCF77 time signals will fail at the same time, but it is possible.
>Therefore was my initial question if it is possible to sync to the PPS
>signal, even if both prefered peers disappear.

So, again I am confused. 
I have
server tick.usask.ca
server ntp.ubc.ca
server 127.127.28.0 minpoll 4 prefer

where 127.127.28.0 is the PPS via shm and this system seems to have no
problems.
So again my confusion may be irrelevant (well my confusion certainly is
irrelevant but my situation may not be.) Or maybe I simply do not recognize
the problem.
And why is the pps not the preferred peer?
(By GPS I assume you mean an NMEA GPS source).

 

>Furthermore the used cesium normale is part of the UTC calculation of
>the BIPM in Paris. So the cesium timings are constantly measured and
>ensured by Circular T: ftp://ftp2.bipm.fr/pub/tai/publication/cirt/cirt.254
>;-)




More information about the questions mailing list