[ntp:questions] Refclock not going sys.peer when there is a pps.peer

simonpg paulgsimons at gmail.com
Wed Feb 18 19:09:24 UTC 2009


I recently added an ATOM refclock based on the PPS output of my
TRUETIME 486-DC with the G2G antenna.  The results are:

# ntpq -c rv -c ass -p
associd=0 status=0115 leap_none, sync_pps, 1 event, clock_sync,
version="ntpd 4.2.5p158 at 1.1809-o Mon Feb  9 04:32:17 UTC 2009 (1)",
processor="i686", system="Linux/2.6.26.simonet", leap=00, stratum=1,
precision=-19, rootdelay=0.000, rootdisp=0.373, refid=PPS,
reftime=cd46d043.8c33cccc  Wed, Feb 18 2009 10:21:55.547,
clock=cd46d049.56564abc  Wed, Feb 18 2009 10:22:01.337, peer=33855,
tc=4, mintc=3, offset=0.063, frequency=-89.057, sys_jitter=0.002,
clk_jitter=0.002, clk_wander=0.001

ind assid status  conf reach auth   condition    last_event      cnt
===========================================================
  1 33855  974a   yes   yes  none   pps.peer    sys_peer       4
  2 33856  941b   yes   yes  none   candidate   clock_alarm  1
  3 33857  941d   yes   yes  none   candidate                       1
  4 33858  9324   yes   yes  none   outlyer       reachable      2
  5 33859  9337   yes   yes  none   outlyer       raate_exceeded  3
  6 33860  9344   yes   yes  none   outlyer       reachable      4
  7 33861  933d   yes   yes  none   outlyer
3
     remote           refid              st t when poll reach
delay   offset  jitter
==============================================================================
oPPS(1)              .PPS.           0 l     6   16  377    0.000
0.063   0.002
+TRUETIME(0)     .TRUE.         0 l   52   64  377    0.000   -6.287
0.053
+clepsydra.dec.c  .GPS.          1 u   51   64  377   53.204
7.740   3.352
-clock.xmission.   .GPS.          1 u   41   64  377   73.440
8.210   0.853
-bigben.cac.wash .USNO.        1 u   32   64  377   40.972    8.035
1.217
-131.107.13.100   .ACTS.         1 u     6   64  377   52.141
19.335   1.796
-clock.sjc.he.ne   .CDMA.        1 u   57   64  377   53.710
8.522   0.623

ntpq> clockv 33856
associd=33856 status=00f0 , 15 events, clk_unspec,
device="Kinemetrics/TrueTime Receiver", timecode="049:18:34:08 ",
poll=10574, noreply=0, badformat=0, baddata=0, fudgetime1=0.000,
stratum=0, refid=TRUE, flags=0
ntpq>

Shouldn't the TRUETIME clock assume the sys.peer condition at some
point?  This configuration has been running for about a week.  What is
the next step?




More information about the questions mailing list