Eugen COCA wrote: > On Apr 7, 10:36 am, [EMAIL PROTECTED] (Hal > Murray) wrote: > >>>I still have a question unanswered: are there any method to pass over >>>a long GPS failure (say for several days or more) - supposing the GPS >>>is the >>>only source of time and the server have a very good (10E-12) PPS >>>source (Rubidium or Caesium) ? >> >>Google for Z3801A. It's basically a box with a very good crystal >>oscillator, a GPS receiver, and a pile of software. If the GPS >>goes down, it coasts on the crystal. >> >>How long you can coast depends upon how good your clock is and/or >>your accuracy requirements. > > > > Hal, > > I have the GPS receiver (many), I have the Rubidium standard (many), > I have a PC with FreeBSD (or any other good OS needed) and, of course, > we all have NTPD. > My questions was how can I make this system to depend as little as > possible on the GPS signal. > > I exposed a solution - by the two "prefer" kewords. This is a good > solution as the drift after > four days was negligible (<1E-6). > > Other opinions on this configuration ? > > >
I don't believe that more than one "prefer" keyword is useful. ISTR reading here that ONLY ONE source can be "preferred". _______________________________________________ questions mailing list [email protected] https://lists.ntp.isc.org/mailman/listinfo/questions
