[ntp:questions] Stick to PPS, even if the prefer server fails

alkopedia at googlemail.com alkopedia at googlemail.com
Tue Mar 24 18:10:04 UTC 2009


On Mar 23, 5:00 am, Dave Hart <daveh... at gmail.com> wrote:
> Dr. Mills,
>
> Thank you for your confirmation.  The original poster should be
> satisfied, assuming their kernel time is being disciplined directly by
> their cesium standard.

What do I have to do to use the kernel discipline?
At the moment I'm using a Linux kernel 2.6.28.1 with LinuxPPS patch
(http://wiki.enneenne.com/index.php/LinuxPPS_support). The PPS signal
is connected via RS232 DCD pin and used by the ntpd refclock_atom
driver.
I don't see the PPSTIME bit in ntptime:

monitor202[~]# ntpq -pn
     remote           refid      st t when poll reach   delay
offset  jitter
==============================================================================
+192.168.12.3    .GPSi.           1 u    8   64  377    1.047
2.127   0.194
 192.168.12.5    .INIT.          16 u    -  128    0    0.000
0.000   0.000
o127.127.22.0    .PPS.            1 l    8   16  377    0.000
-0.027   0.028
 192.168.12.11   192.168.12.3    13 u   10   64  377    0.979
-152.47   0.147

monitor202[~]# ntpq -crv
assID=0 status=21b4 leap_none, sync_atomic/PPS, 11 events, event_peer/
strat_chg,
version="ntpd 4.2.4p5 at 1.1541-o Mon Dec  1 14:26:35 UTC 2008 (1)",
processor="i486", system="Linux/2.6.28.1elan", leap=00, stratum=2,
precision=-16, rootdelay=0.000, rootdispersion=0.351, peer=63523,
refid=PPS(0), reftime=cd739da1.0860c62e  Tue, Mar 24 2009
17:57:53.032,
poll=4, clock=cd739da3.1eedef80  Tue, Mar 24 2009 17:57:55.120,
state=4,
offset=-0.032, frequency=77.428, jitter=0.026, noise=0.092,
stability=0.005, tai=0

monitor202[~]# ntptime
ntp_gettime() returns code 0 (OK)
  time cd739daa.7ecb5320  Tue, Mar 24 2009 17:58:02.495, (.495290872),
  maximum error 5074 us, estimated error 91 us, TAI offset 0
ntp_adjtime() returns code 0 (OK)
  modes 0x0 (),
  offset -30.903 us, frequency 77.428 ppm, interval 1 s,
  maximum error 5074 us, estimated error 91 us,
  status 0x2001 (PLL,NANO),
  time constant 4, precision 0.001 us, tolerance 500 ppm,




More information about the questions mailing list