xntp-alpha thinks my PSTI clock is insane

xntp-alpha thinks my PSTI clock is insane

Post by Jim Bru » Wed, 14 Jul 1993 04:45:15



A few weeks ago, I installed xntp-alpha (May 25/93) on my Sun and got
it working with a PSTI 1020 clock. While I was on vacation, xntp
decided that the PSTI clock was insane and now refuses to synch to it.
Xntp will however synch to the local CPU clock and run at stratum 10
and it will happily serve time to a couple of other systems. (The June
21/93 version of xntp-alpha doesn't work with my clock either.)

If anyone out there is successfully running xntp-alpha with a PSTI
clock, I would appreciate some help. The numbers coming out of ntpq
and xntpdc look reasonable to me (output appended). So why is my clock
insane?

-----
elmo:1> ntpq
ntpq> ass
ind assID status  conf reach auth condition  last_event cnt
===========================================================
  1 47140  9014   yes   yes  none    insane   reachable  1
  2 47141  9614   yes   yes  none  sys.peer   reachable  1
ntpq> pst 47140
status=9014 reach, conf, 1 event, event_reach
srcadr=WWV_PST(0), srcport=123, dstadr=127.0.0.1, dstport=123, keyid=0,
stratum=0, precision=-9, rootdelay=0.00, rootdispersion=0.00,
refid=WWVH, reftime=afec35f4.976c8b42  Mon, Jul 12 1993 15:11:48.591,
delay=    0.00, offset= 4266.88, dispersion=1.60, reach=376, valid=8,
hmode=3, pmode=4, hpoll=6, ppoll=6, leap=00, flash=0x0,
org=afec35f0.5704b000  Mon, Jul 12 1993 15:11:44.339,
rec=afec35f0.5704b000  Mon, Jul 12 1993 15:11:44.339,
xmt=afec35f1.3fce4000  Mon, Jul 12 1993 15:11:45.249,
filtdelay=    0.00    0.00    0.00    0.00    0.00    0.00    0.00    0.00,
filtoffset=4266.94 4266.88 4267.32 4266.82 4266.91 4266.74 4267.30 4267.07,
filterror=    0.75    1.51    2.85    3.36    4.38    5.45    6.47    7.29
ntpq> q
elmo:2> xntpdc
xntpdc> showpeer 127.127.3.0
remote 127.127.3.0, local 127.0.0.1
hmode client, pmode unspec, stratum 0, precision -9
leap 00, refid [WWVH], rootdistance 0.0000, rootdispersion 0.0000
ppoll 6, hpoll 6, keyid 0, version 3, association 47140
valid 8, reach 376, unreach 0, trust 000
timer 38s, flags configured, reference_clock, preferred_peer
reference time:      afec3634.97ae147a  Mon, Jul 12 1993 15:12:52.592
originate timestamp: afec3630.576a9000  Mon, Jul 12 1993 15:12:48.341
receive timestamp:   afec3630.576a9000  Mon, Jul 12 1993 15:12:48.341
transmit timestamp:  afec3631.3ff9e000  Mon, Jul 12 1993 15:12:49.249
filter delay:  0.0000   0.0000   0.0000   0.0000  
               0.0000   0.0000   0.0000   0.0000  
filter offset: 4.26673  4.26694  4.26688  4.26732
               4.26683  4.26692  4.26675  4.26731
filter order:  0        2        1        4      
               3        5        6        7      
bdelay filter: 0.00000  0.00000  0.00000  0.00000
               0.00000  0.00000  0.00000  0.00000
delay 0.0000, estbdelay 0.0000
offset 4.266729, dispersion 0.0010
xntpdc> clkbug 127.127.3.0
clock address:        127.127.3.0
values: 14
           8           0           0          16
         193          14          13          32
         601           5           5          50
  2934835200           0
times: 26
2951493172.359515 93:192:19:12:52.3592951493176.610500 93:192:19:12:56.610
2951493176.348342 93:192:19:12:56.3482951493180.599500 93:192:19:13:00.599
2951493180.369560 93:192:19:13:00.3692951493184.619500 93:192:19:13:04.619
2951493188.344736 93:192:19:13:08.3442951493192.595500 93:192:19:13:12.595
2951493192.363433 93:192:19:13:12.3632951493196.615500 93:192:19:13:16.615
2951493200.341163 93:192:19:13:20.3412951493204.592500 93:192:19:13:24.592
2951493204.359801 93:192:19:13:24.3592951493208.611500 93:192:19:13:28.611
2951493208.348719 93:192:19:13:28.3482951493212.601500 93:192:19:13:32.601
2951493152.355567 93:192:19:12:32.3552951493156.607500 93:192:19:12:36.607
2951493156.344951 93:192:19:12:36.3442951493160.595500 93:192:19:12:40.595
2951493164.351962 93:192:19:12:44.3512951493168.603500 93:192:19:12:48.603
2951493168.341470 93:192:19:12:48.3412951493172.592500 93:192:19:12:52.592
2951493208.348719 93:192:19:13:28.3482951493212.601500 93:192:19:13:32.601

--

Ontario Hydro Research Division     Voice: (416) 207-6733
800 Kipling Ave, KR236              Fax: (416) 207-6216
Toronto, Ontario, CANADA, M8Z5S4

 
 
 

1. xntp-alpha and PSTI clock insanity

A few days ago, I asked for some help with xntp-alpha and my PSTI
clock. In short, xntpd thought that my clock was insane even though it
looked good to me. With a bit of help and some digging in the source
code, it appears that the clock was failing a test in ntp_proto.c. The
reference timestamp was about 5 seconds faster than the originate
timestamp so xntpd concluded that the clock was broken (select finds
broken).

Now for the 'interesting' part. I found that my CPU clock was running
around five seconds slow at the time. (I was using this as a stratum 9
backup.) I used the date command to set the CPU clock several seconds
fast. After this, xntp decided that the PSTI clock was OK and
synchronized to it. I assume that the slowness of the CPU clock caused
xntp to think that my PSTI was insane.

Questions:

1. How are the reference and originate timestamps determined? Could
they be based on different ideas of time?

2. Has anyone else noticed this behavior? I presume that this wouldn't
happen normally because xntp would keep things synchronized. However,
I found that there were several power outages here so the PSTI might
have gone away long enough for things to drift.

--

Ontario Hydro Research Division     Voice: (416) 207-6733
800 Kipling Ave, KR236              Fax: (416) 207-6216
Toronto, Ontario, CANADA, M8Z5S4

2. Removing control

3. Help needed with xntp, Sun and PSTI clock

4. Returning array from object implementation

5. PSTI clock problem (was: Local clock bug?)

6. Go directly to the record that must be updated?

7. Patches for xntp-alpha (June 21, 1993) and AT&T StarServer E

8. Do X10 Signals Leave the home?

9. [jones: Re: Bugs in xntp-alpha distribution(Jul 22, 1993)]

10. Solaris 2.2 and xntp-alpha: Ready?

11. Bugs in xntp-alpha distribution(Jul 22, 1993)

12. xntp-alpha and Solaris 2.2

13. breakage in xntp-alpha