[ntp:questions] Re: SHM refclock not being selected
davids at webmaster.com
Mon Mar 14 19:06:48 UTC 2005
"John Ackermann N8UR" <jra at febo.com> wrote in message
news:mailman.23.1110723967.576.questions at lists.ntp.isc.org...
> Here's my configuration file, with logging stuff, etc., trimmed for
> # /etc/ntp.conf, configuration for ntpd
> # Z3801A
> server 127.127.26.0 ttl 1
> fudge 127.127.26.0 time1 -0.967
> fudge 127.127.26.0 refid GPS
> # PPS via linux_shm_clock
> server 127.127.28.0
> fudge 127.127.28.0 refid PPS
> server tock.febo.com
> server time-c.timefreq.bldrdoc.gov
> server ntp1.usno.navy.mil
> server ntp2.usno.navy.mil
> That's it...
You are sure there are absolutely *no* 'restrict' lines?
>> Now that's strange. Does this machine have one CPU or more than one?
>> What type?
> It's an Athlon 2200 or thereabouts. When I was running a 2.4 kernel, I
> had no problems but the change to 2.6 (necessitated by some other software
> that is the machine's primary reason for being) really screwed things up.
I'm pretty baffled. The kernel issue might have something to do with
power management, but I doubt that has anything to do with the clock
I can't think of any reason why NTP would select a clock at a higher
stratum when it has clearly usable references at a lower stratum.
More information about the questions