On Tue, Oct 17, 2006 at 03:35:09PM -0500, Seth Goodman wrote: > Roberto C. Sanchez <roberto@connexer.com> wrote on Tuesday, > October 17, 2006 2:25 PM -0500: > > > $ ntpq -p > > remote refid st t when poll reach delay offset jitter > > ======================================================================== > > yauco.connexer. .INIT. 16 u - 1024 0 0.000 0.000 4000.00 > > maracaibo.conne .INIT. 16 u - 1024 0 0.000 0.000 4000.00 > > It looks like you can't reach your servers, or you reach it and then discard > it (ntp determines it is a 'falseticker'). The first character on each > server line is a space, which indicates its status as a server is 'reject'. > Also, both servers show up as stratum 16, which is not reasonable since each > of the two servers was configured to use three low stratum servers. If > working, you would see a '+' in the first column, the 'stratum' column would > show 2 or 3, the 'when' column would show a number less than the poll > column, > the 'reach' column would typically show 377, and the 'delay', 'offset' and > 'jitter' columns would show non-zero values. > > When a machine can't reach any of its designated servers, it uses the local > hardware clock, so that's probably why it's drifting. > > Once you resolve the reachability issue, you might check whether a drift > file is declared in ntp.conf. The drift calculation takes about a day to > stabilize and with no drift file declared, ntp doesn't write this out at > shutdown and must start over each time the machines reboots. To do the > peering that Henrique suggested, you declare the other server as a peer. > Here's what ntp.conf would look like for yauco.connexer.com: > > driftfile /var/lib/ntp/ntp.drift > > server ntp2.usno.navy.mil > server ntp-1.vt.edu > server ntp-2.vt.edu > > peer maracaibo.connexer.com > OK. This, in conjunction with fixing my broken "restrict" statements did the trick. Basically, when I set up ntp years ago on my machines, I was not running any local ntp server, so they all hit the public ntp server and I basically had my machines set so that they would not synchronize each other. Whem I setup the ntp servers on my network which were to synchronize from the public ntp servers and then provide ntp service to my other servers and workstation, I just copied an ntp.conf from one of the workstations. Of course, this made it so that the clients wouldn't synchronize since I had use the restric keyword to prevent it. A bit of reading through the documentation let me figure out why what I was doing with that was wrong. So bascically, my workstations and servers have not been getting the benefit of NTP for nearly a year now. It's good to be back on track. Regards, -Roberto -- Roberto C. Sanchez http://people.connexer.com/~roberto http://www.connexer.com
Attachment:
signature.asc
Description: Digital signature