[ntp:questions] rdate vs ntpd -q

bombjack bombjack99 at gmail.com
Tue Mar 22 20:34:35 UTC 2011


On 22 mar, 18:42, unruh <un... at wormhole.physics.ubc.ca> wrote:
> On 2011-03-22, bombjack <bombjac... at gmail.com> wrote:
>
> > Hi,
> > I am currently working on a commercial system which uses rdate for
> > roughly setting time when booting/restarting. Later ntpd is started
> > and used to keep time accurate. My question: Is it possible to get rid
> > of rdate in favour of "ntpd -q"? Any risks involved? ntpd starting
> > later and/or needs more time to sync than rdate etc?
>
> ??? why would you use rdate? Yes, ntp -q would do all rdate does and
> with greater accuracy. rdate does not sync. ntpd -q does not sync. Both
> determine the time from elsewhere and set the system clock to that time.
> Boom. But you can also start up ntpd and ask it to set the system time
> immediately. It may then go on a big wander (10s of ms) while it hunts
> for the right drift rate compensation.
> ntpd -q and rdate will both do absolutely nothing about the drift rate.
> ntpd -g and the normal operation of ntp ( it steps the clock if it is
> more than 128ms off) should do what you want anyway.

Hehe, that's what I am trying to do, i.e. get rid of rdate. The system
I work with is rather old and has used rdate since forever. I am
merely doing maintenance work and want to exchange rdate for ntpd.
That's what its all about. ms and drift files are of no concern at the
boot stage. If I can get the hw-clock within ms, I'm home free.

Thanks for the info! I was suspecting an answer like this, but wanted
to check if there are any pitfalls I might have overlooked. I will
have a test with "ntpd -g -q" and take it from there.

//Fredrik




More information about the questions mailing list