Re: NTP with only GPS/PPS flags PPS as falseticker

[Greg Troxel]

"Nick Burkitt" <address@hidden> writes:

> I've fiddled with the value of the time1 option on the shared memory
> driver refclock line (refclock shm unit 0 refid GPS minpoll 3 maxpoll
> 4 time1 0.080 flag4 1), and it seems to make a slight
> difference. Occasionally I see a '+' next to SHM(0) in nptmon when
> disconnected from the network, but only briefly. Although the PPS
> offset no longer seems to drift...

So you are at 6.5ms different from "GPS" to "PPS".  Not really that bad
in the historical view.

(It would be good if you could send mail with output that is not munged,
but I know not everyone uses emacs!.)

>>   Read about minsane.  If you want to sync to just GPS, then you need to
>>   accept a single source.  ntpd by default does not do this.
> The documentation (https://docs.ntpsec.org/latest/miscopt.html) claims
> that the default value for minsane is 1

So set it anyway and check that assumption.

> I'll try my luck on the ntpsec mailing list.

Sounds good - the real questions are for ntpsec now.

Let us know if you make progress.