[ntp:questions] improving ntpd performance on Windows

Dave Hart davehart at gmail.com
Fri Feb 27 17:23:49 UTC 2009


On Feb 27, 1:15 am, I wrote:
> As a test I fiddled with the set of internet servers carefully to
> minimize clockhopping to generally cause 1 msec or less steps and
> overall manage to stay within +/- msec of the changing sources.  With

Sorry I meant "stay within +/- 2 msec" regarding that marginally
stable LAN source.

Looking at a machine on a different network with a refclock, I noticed
a +/- 100 usec sawtooth with a 4 hour period:

http://davehart.net/ntp/refclock/20090227-refclock-dirty-sawtooth.jpg

Cheers,
Dave Hart




More information about the questions mailing list