[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [gpsd-dev] Draft Time Service HOWTO is beginning to approach final f
Re: [gpsd-dev] Draft Time Service HOWTO is beginning to approach final form
Thu, 24 Oct 2013 02:56:19 +0000
"Gary E. Miller" writes:
> On Thu, 24 Oct 2013 01:39:38 +0000
> Harlan Stenn <address@hidden> wrote:
> > > 1. ntpd has a built in hold timer at startup before it does any
> > > grooming of the clock. It defaults to 300 seconds:
> > > http://www.eecis.udel.edu/~mills/ntp/html/clock.html
> > > http://www.eecis.udel.edu/~mills/ntp/html/miscopt.html#tinker
> > > That's a nice hard minimum time to wait for the HOWTO.
> > That's a choice if one is only looking to sync with a local refclock
> > that may be in cold-start.
> > What would be better is to use this as a means to motivate GPS and
> > other refclock vendors to do a better job reporting "state" -
> > precision, error, etc.
> gpsd used to report expected precision to ntpd but it never did anything.
> So it was removed. see ntpshm.c around line 394
> If there is a way to give it, we'll give it.
> > If one has a good drift file and the upstreams are in good shape, ntpd
> > can do a restart and achieve full sync in about 11 seconds' time.
> That is a LOT of assumptions! And pretty unbelieable since the default
> poll time is 64 seconds.
iburst sends packets a 2 second intervals on the initial start and 5
responses are needed to hit critical mass. So about 11 seconds' time.
If one doesn't use iburst that turns in to about 5 minutes, as the polls
are sent every 64 seconds (by default).