[LON-CAPA-admin] ntp

Martin Siegert siegert at sfu.ca
Tue Mar 25 13:08:27 EST 2003


The problem is in all likelyhood that your machine synchronizes to itself
(localhost). You must explicitly tell it not to do that. Here is the
/etc/ntp.conf file that I am using on dalton.chem.sfu.ca:

===</etc/ntp.conf example>=============================================
driftfile /etc/ntp/ntp.drift
server  ntp.sfu.ca      # 142.58.103.1  whistler.sfu.ca
server  seymour.sfu.ca  # 142.58.103.2
server  morgoth.sfu.ca  # 142.58.101.12
server  rm-rstar.sfu.ca # 142.58.120.21
#
# Setup default restrictions
#
restrict default ignore
#
# Servers
#
restrict 142.58.103.1 noquery		# whistler
restrict 142.58.103.2 noquery		# seymour
restrict 142.58.110.12 noquery		# morgoth
restrict 142.58.120.21 noquery		# rm-rstar
#
# Answer our own requests but don't synchronise
#
restrict 127.0.0.1 notrust
========================================================================

The critical line is the last line. You should always check with
"ntpq -p" whether ntp is working correctly. On dalton this gives:

# $ ntpq -p
     remote           refid      st t when poll reach   delay   offset  jitter
==============================================================================
*whistler.sfu.ca time.nist.gov    2 u  530 1024  377    0.708    3.443   3.139
+seymour.sfu.ca  whistler.sfu.ca  3 u  444 1024  377    0.602   -9.743   0.482
 morgoth.sfu.ca  0.0.0.0         16 u    - 1024    0    0.000    0.000 4000.00
+rm-rstar.sfu.ca usno.pa-x.dec.c  2 u  490 1024  377    0.702  -12.283  10.777

The "morgoth.sfu.ca" line is typical for a ntp server that is not working.
If you have only such lines (or just a localhost line), you'll see the
effects that you are describing.

Martin

On Tue, Mar 25, 2003 at 08:03:09AM -0500, Gerd Kortemeyer wrote:
> We wish we knew. We have been struggling with the RedHat setup of NTP 
> for quite a while.
> 
> On Tuesday, March 25, 2003, at 07:55  AM, H. K. Ng wrote:
> 
> >Hi,
> >
> >One of my lon-capa servers time is off even though the ntpd is 
> >running. When I restart it again, it says that it in sync with our 
> >official clock but is almost 5 minutes off - i.e., it does not reset 
> >to the correct time. The ntp.conf and step.tickers are identical for 
> >the two machines. Any suggestions why?
> >
> >Regards,
> >hk
> >
> >
> >_______________________________________________
> >LON-CAPA-admin mailing list
> >LON-CAPA-admin at mail.lon-capa.org
> >http://mail.lon-capa.org/mailman/listinfo/lon-capa-admin
> 
> _______________________________________________
> LON-CAPA-admin mailing list
> LON-CAPA-admin at mail.lon-capa.org
> http://mail.lon-capa.org/mailman/listinfo/lon-capa-admin



More information about the LON-CAPA-admin mailing list