Try ntpq -p and see what it says.  Should have something like this:

     remote           refid      st t when poll reach   delay   offset  jitter
==============================================================================
*yourserver.yourdomain.com ntp-0.cso.uiuc.  3 u  417 1024  377    0.555
4.640 1.773

word wrap is messing up this display but you can get the idea.

most important thing in all that is the reach info.  desired number
there is 377, but anything above 0 means it's connecting a little to
your outside time source.

the other important thing is that the timeserver you have chosen to
connect to allows connections to it.  Some don't any more due to the
great number of people connecting to them.  try ntp-0.cso.uiuc.edu, it's
one I'm using at the moment.  

If you still can't get it going, feel free to E-Mail me your ntp.conf
file and I'll give it a look.  ntp can be a battle to get going at first,
but I did finally get it right and it's doing a great job now.


On Sat, 1 Dec 2001 14:06:03 -0800 (PST)
"Charlie Bebber" <[EMAIL PROTECTED]> wrote:

Charlie> 
Charlie> I had this working many moons ago, but with new upgrades/installs, it's all
Charlie> been lost and I can't remember what I did (but I know it's not that hard --
Charlie> that's why it's driving me crazy).
Charlie> 
Charlie> So here's what I've got going on:
Charlie> 
Charlie> - ntp installed on all of my machines
Charlie> - ntp configured on fileserver to sync to external time server (doesn't stay
Charlie> synced, however).
Charlie> - ntp configured on remaining nodes to sync with fileserver who is supposed
Charlie> to  allow that
Charlie> 
Charlie> But what's happening is that the ntp on the fileserver will start up, sync
Charlie> with the external time server, but down the line, the time begins to drift
Charlie> and according to my syslog, there were never any more attempts to keep time
Charlie> synced  (ntp-4.1.0-1mdk installed on every one of them).
Charlie> 
Charlie> And on the local clients, in my /etc/ntp.conf, I've got 10.1.1.3 (the IP for
Charlie> the fileserver) set for the server.  That never worked so I stuck it in my
Charlie> /etc/ntp/step-tickers and still, no love.  Here's what I get:
Charlie> 
Charlie> ntpdate[13387]: no server suitable for synchronization found
Charlie> 
Charlie> I portscanned the fileserver and 123 isn't even open (shouldn't it be
Charlie> listening on that port?).  When I do a 'ps auxw |grep ntp' on the
Charlie> fileserver, all I've got is 'ntp -A'.  Should there be anything else in
Charlie> order to allow other nodes to sync?
Charlie> 
Charlie> So what am I missing here?  I'm all out of ideas.
Charlie> 
Charlie> Thanks in advance,
Charlie> 
Charlie> -Charlie
Charlie> -- 
Charlie> GPG Key fingerprint = 4F36 EC4F 2F2C 5F59 9690  09E5 4C0F 9DB0 8623 53CE
Charlie> 
Charlie> 
Charlie> 


--
Brian - [EMAIL PROTECTED]
My Home Page: http://www.brimac.com/~brianmac
Fine Photos: http://www.brimacphotography.com
Art for Sale: http://www.artbrowser.com
Classified Advertising: http://www.sellit2000.com


The only 15 letter word that can be spelled without repeating a letter 
is uncopyrightable.


Want to buy your Pack or Services from MandrakeSoft? 
Go to http://www.mandrakestore.com

Reply via email to