[LinuxPPS] still the strangeness

[Bernhard Schiffner]

Am Dienstag, 19. Dezember 2006 16:04 schrieb Udo van den Heuvel:
> Hello,
>
> Today I upgraded to ntp 4.2.2p4. No change in behaviour:
>
...
> Any comments? Any ideas?
> Udo

Step by step:
(general)

1.) give me a uname -a for information please.

2.) comment out all peers in ntp.conf

3.) put in _one_  you can really trust (for me it is ntp1.ptb.de) 

4. ) restart ntpd
 (be aware that ntpd with 2.6.18+ kernels take up to 4 hours to become stable)

5.) give me the ping-time to this peer for information please.

6.) report please what's on with your other peers with ntpdate -b -q x.y.z.a 
&& ntpq -c pe

-------------------
(pps general)
(still not as clocksource in ntp.conf)

7.) read from your pps source (by minicom)

8.) cat /proc/pps/0*/* to see the timing of assert and clear events

9.) one slope of the pps_signals should be within 10 ms to the full second, 
these 10 ms result from asymmetric connection (modem, adsl)

10.) decide from this  wich slope is yours (about 200 ms from the other)

11.) if you don't find your pps-signal within 10 ms to your ntp-peer check the  
reported satellites used:
$GPGSA,A,3,04,05,,09,12,,,24,,,,,2.5,1.3,2.1*39
(normaly >=4)


-------
(use of gps clocksource in ntp.conf)
later

(use of pps_signal)
later

Bernhard