NTP configuration troubles

[pkoning]

I'm trying to get NTP running on my 9.2 system. The config file looks like this:

server 192.168.10.187
broadcast 192.168.10.255
driftfile
logconfig +sysevents +syncstatus

That server is a GPS clock server on the LAN, and NTP can indeed sync with it (ntpd -G -q works). But when I use this config and start the NTP process the startup says it worked, but there isn't any indication in the log that it's doing anything. Nor do I see the expected broadcasts. Is something else needed? The documentation is not overly clear...

--paul

[martinv]

Paul,

I just configured NTP on my x86 V9.2-1 system, and it worked right away.

I copied SYS$SPECIFIC:[TCPIP$NTP]TCPIP$NTP.TEMPLATE to .CONF and edited it
- comment out the peer and server example statements
- insert my local NTP server
- uncomment the backup NTP server statements
After that it looks like this (without the comments)

Code: Select all

driftfile SYS$SPECIFIC:[TCPIP$NTP]TCPIP$NTP.DRIFT
server my.local.ntp.server prefer
server 127.127.1.0
fudge 127.127.1.0 stratum 8

I don't see a logconfig statement in the documentation, and no ntpd command either.
Correction: I found logconfig now.

After starting NTP, I monitor it with NTPQ:

Code: Select all

$ ntpq

ntpq> hostnames no
ntpq> peers
     remote           refid      st t when poll reach   delay   offset  jitter
==============================================================================
 127.127.1.0     .LOCL.           8 l  645   64    0    0.000   +0.000   7.812
*w.x.y.z         a.b.c.d          2 u   40   64  377    7.812   -2.685   7.812
ntpq> assoc
ind assid status  conf reach auth condition  last_event cnt
===========================================================
  1 28778  8013   yes    no  none    reject unreachable  1
  2 28779  961a   yes   yes  none  sys.peer    sys_peer  1
ntpq> readvar
associd=0 status=0615 leap_none, sync_ntp, 1 event, clock_sync,
version="ntpd 4.2.8p15@1.3728 Fri Sep 22 07:00:58 UTC 2020 (2)",
processor="X86VMS", system="OpenVMS/V9.2-1", leap=00, stratum=3,
precision=-7, rootdelay=20.538, rootdisp=28.920, refid=w.x.y.z
reftime=e8e3312a.5d429475  Wed, Oct 25 2023  6:16:10.364,
clock=e8e3315e.69ed7bfa  Wed, Oct 25 2023  6:17:02.413, peer=28779, tc=6,
mintc=3, offset=-2.684771, frequency=-5.154, sys_jitter=7.812500,
clk_jitter=7.960, clk_wander=0.007
ntpq>

For completeness

Code: Select all

$ tcpip show version

  VSI TCP/IP Services for OpenVMS x86_64 Version 66.0
  on a VMware, Inc. VMware7,1 running OpenVMS V9.2-1

HTH,
Martin

[babydr]

Please be -very- careful of what you implement in the 127.127.0.0./16 arena in NTP .
Please see ... For the list of what each number of the third octet is used for ,

[url]https://www.eecis.udel.edu/%7Emills/ntp ... .html#list[/url]

Hth , JimL

[martinv]

The lines containing 127.127.1.0 are straight out of the template file, and - as the VMS NTP implementation does not support orphan mode - are the best we have. They're not really necessary in my setup, though.

Added in 1 hour 17 minutes 59 seconds:
One thing I noticed in my NTP experiments: when I shutdown and start NTP, the process TCPIP$NTP_1 runs at 100% CPU load for some minutes, which - at priority 10 - makes interactive use of the system very slow. I wonder if that's only my setup or if it's a general behaviour.