[ntp:questions] Second attempt at GPS-18 based NTP server

unruh unruh at invalid.ca
Fri Feb 3 05:47:15 UTC 2012


On 2012-02-03, David Lord <snews at lordynet.org> wrote:
> David Woolley wrote:
>> David Lord wrote:
>> 
>>>
>>> I've just added the "minpoll 4 maxpoll 4" back to my ntp.conf
>>> as it could reduce the 30 usec offset blips I get when logs are
>>> rotated.
>> 
>> That's almost certainly totally the wrong thing to do.  If you have 
>> short periods of increased interrupt latency or network latency, you 
>> want a long poll interval so that the samples with the high offsets (and 
>> delays) are ignored.
>> 
>> Forcing a short interval will make ntpd track the temporary errors in 
>> the time measurements.
>> 
>> Offset is not a direct measure of the timekeeping error.
>
> I already know that the system can't correct for temperature
> changes, as with poll interval 10, and no PPS, the offset will
> just carry on increasing until stepped or until the temperature
> change is reversed. If the cause of these blips is down to
> temperature I'd expect the shorter poll interval to be better.
> I'll leave it at "4" for a couple of days then have another few
> days with poll increased to "8".
>
>
> The blips are nearly always present, 15u up to 40u.

Just looking at my system, and I get blips of up to 20 in the difference
in offset from one second to the next as well. No idea why such huge
differences.

>
>
> David



More information about the questions mailing list