[ntp:questions] Change reference clock soon afterDCF-signalislost

unruh unruh at wormhole.physics.ubc.ca
Fri Jul 16 15:43:14 UTC 2010

On 2010-07-16, Matuschka, Sebastian <Sebastian.Matuschka at gcd-solutions.de> wrote:
>>I think you mean that you have a 1PPS source that is synchronized to
>>DCF-77 but does not send the actual received DCF-77 signal.
>>In that case it should be OK to use it to increment a clock.
> No, it really is a DCF-77 signal.

Uh, why don't you tell us exactly what you have. Why in the world would
any local source be using a DCF77 signal?

>>However, when you have expectations that you can control what source
>>is used to sync your clock, and how quickly that switches, you cannot
>>use ntpd.  It does not offer that control.
> I don't want to control what source is used, just how fast it is switched when the signal is lost.
> minpoll and maxpoll seem to control that, and i guess i just need smaller maxpoll than the one that is allowed.

And exactly how would it know that the signal is  lost and not just
temproarily missing. Is one dropped packet enough?(that would be a
disaster for any real DCF77 source). 

> I have set minpoll to 4 and maxpoll to 5. After the signal is lost it takes 3 to 4 minutes to switch to another source.


> Why does it take so long when maxpoll is 5 (32 seconds)? How many times must a poll fail until the source is expected to be unrecoverrable for now?

Because it waits a while to see if the loss is temporary. 
Probably 8 times. 

> Best Regards
> Sebastian

More information about the questions mailing list