[Slackdocs] Proofreaders wanted for NTP-HOWTO

x y x77 at shaw.ca
Sat Sep 8 22:22:45 CEST 2012

On 09/08/2012 12:14 PM, Niki Kovacs wrote:
> x y a écrit :
>> Watch that. If you only sync once, clock drift (or a bad battery causing
>> clock drift) on some machines will put them out as much as several
>> minutes in 24 hours. Things like network encryption depend on accurate
>> time on all machines or they start slowing down and throwing errors and
>> not connecting. It can be a big problem to diagnose.
> I think there's some misunderstanding here.
> 1) The initial 'ntpdate' sets the clock right manually, the first time.
> 2) Once the network is setup as described in the HOWTO, machines are 
> *always* in sync. No need for any cronjob, stanza in rc.local or the 
> likes. Just point ntp.conf to your server, activate and start rc.ntpd 
> and you're done.
> You can actually check this out by yourself (which I do when I setup a 
> LAN).
> Once everything's configured, run this command on all the clients:
> # watch ntpq -p
> And then sit back and see how all your clients sync one by one to the 
> server... and stay in sync.
> cheers,
> Niki
Oop, I see what I did there. Yes, you are using the pools for the sync, 
which is fine and what you have there is correct.

Sorry, next time I'll read the post with my good eye... ;)

More information about the Slackdocs mailing list