[ntp:questions] Change reference clock soon afterDCF-signalislost

Rob nomail at example.com
Fri Jul 16 11:57:39 UTC 2010

Matuschka, Sebastian <Sebastian.Matuschka at gcd-solutions.de> wrote:
>>You are also still hiding stuff and expecting us to guess.
> That's absolutely not my intention. I just want to write only the things that are relevant, so people don't have to read so much they decide to not read it at all.
> What could be of intereset for you? What can i explain to help you helping me?
> Sorry if anything is missing, i really don't want that.

There are regular postings of questions that are similar to yours.
(i.e. somebody wants ntpd to act quicky on events and in a controllable way)

As I see it, ntpd simply isn't like that.
It has a mind of its own, and it acts slowly and patiently.

When you want tight control over what source is used for the clock and
how quickly it decides to use another clock, ntpd isn't for you.

What fails from your explanations is why you want that behaviour.
The purpose of ntpd is to have an accurate clock within the limitations
of available sources, but you seem to have different requirements.

More information about the questions mailing list