[ntp:hackers] Re: NTP and leap-seconds

David J Taylor david-taylor at blueyonder.co.uk
Mon Jul 3 06:40:47 UTC 2006


David L. Mills wrote:
> David,
>
> This is an interesting discussion. Consider the case where the NIST
> advertises a DUT1 change or a DST update or a leapsecond change. This
> is no a Byzantine agenda as the selection algorithm is designed to
> handle, but an arbitrary bit decision. See the WWV/H driver for an
> excursion in that direction. The algorithm is properly descirbed as
> maximum-likelihood from a mathematical point of analysis. Are we
> prepared to go to those extremes>
>
> Dave

Dave,

I can't comment on the complexities or otherwise of a different algorithm. 
It does seem to me, though, that the present leap-second algorithm where, 
apparently, a single bad low stratum server can poison a whole lot of 
higher-stratum servers goes against the general design principles of NTP 
of seeking only the best and generally agreed time.

Hence my suggestion of majority voting rather than accepting a single 
server.  I am thinking more of the mass of clients rather than the fewer 
stratum-1 servers.

Cheers,
David 




More information about the hackers mailing list