[ntp:questions] Re: How to handle leap second condition correctly?

Harlan Stenn stenn at ntp.isc.org
Sat Jun 11 09:29:25 UTC 2005


>> An SNTP server should not ordinarily be telling ANYBODY what time it is.

> An SNTP server that isn't telling someone the time is a total waste of
> resources.  Telling clients the time is the only reason for an SNTP
> server to exist!

I pretty much agree with you.  In most cases, there is no reason to run
SNTP as a server, and doing so is a waste of resources.

Telling clients the time is the job of an NTP-based server, not an
SNTP-based server.  Usually.

> Are you possibly trying to say that an SNTP client should not also act
> as a server?

Yes.

> Whilst true, this is only a SHOULD NOT, not a MUST NOT,
> and Windows violates this rule anyway, which means it has become 
> unenforceable (ignoring that Windows has more serious protocol
> violations).

And there are more cockroaches on the planet than people, but that does not
make them a higher life form.

> In any case, the original article seems to have meant the server response
> to an SNTP client request, not really that the server had to be an SNTP
> server.

The operative word there is "seems".

> Even if it were an SNTP server, there is nothing to indicate
> that it isn't directly connected to a proper reference clock.

Yes, and there is nothing to indicate that if it is an SNTP-based server
that it *is* connected to a proper refclock, either, or that the time from
the refclock is "sane".

We basically need more information, including knowing exactly what is
expected as far as accurate timekeeping is concerned.

H



More information about the questions mailing list