Mailinglist Archive: opensuse (1002 mails)

[Per Jessen <per@xxxxxxxxxxxx>]

Roger Oberholtzer wrote:

> On Fri, 2011-07-15 at 17:04 +0200, Per Jessen wrote:
> > Roger Oberholtzer wrote:
>
> > I may not have understood all of your problem, but without a frequent
> > connection to an accurate time-source, any system will have to rely
> > on a local clock for periods of time.  AFAICT, both chrony and ntp
> > know how to determine the local oscillator drift and massage the
> > clock such that the local clock remains fairly accurate.  If you need
> > a higher degree of accuracy without a more accurate external time
> > source, you have to improve the local time source - temperature
> > controlled xtal oscillator etc.
>
> This massaging does seem to happen when ntp is running and after it
> has a server. But I am not convinced that anything is done when ntp
> starts and before it gets in contact with a server. 

If properly configured ntp will _always_ have a server - the local
clock. The drift of the local clock oscillator will have been
determined, so ntp will use that to keep massaging the clock until a
better source becomes available. 

> And that is what chrony provides. A useful estimate of time before the
> first server is found.

I don't know anything about chrony, so I have to wonder what information
it has that ntp does not? 


-- 
Per Jessen, Zürich (21.4°C)

--
To unsubscribe, e-mail: opensuse+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: opensuse+help@xxxxxxxxxxxx