[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

OT NTP [Re: how can I not have to spend 40 minutes _waiting_ in the morn

From: Todd Denniston
Subject: OT NTP [Re: how can I not have to spend 40 minutes _waiting_ in the morning for each of 40 updates?]
Date: Fri, 10 Oct 2008 11:21:34 -0400
User-agent: Thunderbird (X11/20080707)

chaitu wrote, On 10/09/2008 04:17 PM:
Hi Todd/Arthur,

Thanks for the insights. Unfortunately for me, I am that typical
"young configuration manager" that software companies can afford
easily when they dont have the patience or money to look for an
expert. But with sound advice I intend to learn quickly and do my job
well. Thanks a lot for your help especially!!

Glad to be of help.

The exact background of the problem is here (I hope this answers your

1. All 40 are different operating systems....well technically they are
mostly Linux
3. The cvs update command is not part of the build process, so the
I do
observe the problem of the client boxes losing time daily........these
linux boxes are connected to a Windows NT domain and for keeping
correct time I have configured the NTPD on them by pointing to some
servers in But in spite of that the boxes lose time's a nuisance I could not fix in the last 2 yrs (but
that's for another thread in another forum, I guess). Does it sound
ominous to you?


A reasonably correct ntpd setup would be to have 2 or 3 of your machines peering together to follow reasonable sources such as,,, or [1]. and then all of your other machines should be following the 2 or 3 machines that are following the outside world. Also to _one_ of the server lines on _one_ of the outside following machines I would add
maxpoll 12 burst
to the end of the server line, which lets it not hit the sever as often (only once every 4096 seconds), but when it does it hits it 8 (times in a minute) to get a really good reading. This probably still violates the 'Rules of Engagement'[2] but if you can't get a stable clock otherwise, something has to be done. Of course if you had a bit of money to spend and some time to set it up, you could have your own reference clock[3] to go along with what you get from the world[1].

BTW I have been assuming you are running all your boxes on real hardware, because it is a known problem with vmware (other VMs may be affected, but I have had experience with vmware) that the hosted platform will not be able to keep good time. If you are on vmware, periodic(cron) use of `hwclock hctosys` in the VM hosted OS, with the bare metal OS using NTP is the better option.

[1] a group of folks have banded together to help try to have a good time around the world :)
see is an pool for India, if I understand
also see:



Todd Denniston
Crane Division, Naval Surface Warfare Center (NSWC Crane)
Harnessing the Power of Technology for the Warfighter

reply via email to

[Prev in Thread] Current Thread [Next in Thread]