[ntp:questions] Re: ntpq times out if NMEA refclock configured?

Richard B. Gilbert rgilbert88 at comcast.net
Sun May 14 02:53:41 UTC 2006


R Jenkins wrote:

> Hi,
> 
> I'm trying to add a GPS refclock to my server.
> After total failure with a basic Trimble TSIP output GPS plus the parse 
> clock, I'm now using a Garmin GPS25 and the NMEA refclock.
> 
> I'm getting a very strange effect:
> 
> If the NMEA refclock is enabled in my ntp.conf, ntpq stops working, it just 
> times out.
> ntptime still works any gives a something like reasonable display.
> 
> I have three other servers listed in ntp.conf & simply commenting out the 
> NMEA refclock lines allows ntpq to work again.
> 
> I've tried commenting out all the restrict lines, this does not change the 
> effect.
> 
> This is what I'm adding in ntp.conf - I've also tried it with & without the 
> 127.127.22.0 PPS driver & 'enable pps' command.
> 
> # NMEA Clock using Garmin GPS
> server 127.127.20.0 prefer
> fudge 127.127.20.0 refid GPS flag3 1 time1 0.042
> 
> ntptime produces this:
> 
> ntp_gettime() returns code 0 (OK)
>   time c810bfbb.0cb40000  Sat, May 13 2006 21:27:39.049, (.049622),
>   maximum error 192016 us, estimated error 16 us
> ntp_adjtime() returns code 0 (OK)
>   modes 0x0 (),
>   offset 0.000 us, frequency -495.911 ppm, interval 4 s,
>   maximum error 192016 us, estimated error 16 us,
>   status 0x1 (PLL),
>   time constant 0, precision 1.000 us, tolerance 496 ppm,
>   pps frequency -495.911 ppm, stability 0.000 ppm, jitter 0.000 us,
>   intervals 0, jitter exceeded 0, stability exceeded 0, errors 0.
> 
> but ntpq does this after about 10 seconds:
>  ntpq -c peers
> localhost.localdomain: timed out, nothing received
> ***Request timed out
> 
> I've previously added the ppskit-lite patch to the kernel, which is 2.6.16.9 
> on Centos 4.3 x86-64, Athlon 64 CPU.
> I have udev configured to link /dev/gps0 to /dev/ttyS0 which I believe is 
> what the NMEA refclock expects (& also to /dev/pps0 for the pps clock).
> The GPS is on ttyS0 with the PPS signal converted to +/- 12V on pin 1.
> 
> I've tried the standard Centos RPM for NTP & I'm now using one built on the 
> machine from the Redhat source rpm (ntp-4.2.0.a.20040617-4.src.rpm).
> This seems to be configured as standard to enable all refclocks & it appears 
> to be recognising the kernel PPS capability during the configure stage.
> 
> Any ideas appreciated!
> 
> Robert Jenkins.
> 
> 
> 

After rereading a little more carefully, I notice that your frequency 
correct of -495.9 PPM is on the ragged edge of the 500 PPM limit.  It is 
unusual for a clock to have a freqency error this large; most are below 
50 PPM in absolute value.

Does your system have a kernel parameter called "HZ"?  Is it set to a 
value greater than 100?  I believe I have seen references to values of 
both 250 and 1000; neither value works well with NTPD.  The system seems 
to lose clock interrupts when HZ is greater than 100.  YMMV but if you 
are not using 100, give it a try.




More information about the questions mailing list