On 10/07/2014 23:51, David Woolley wrote:
On 10/07/14 09:23, David Taylor wrote:
although Windows can be kept within a millisecond,
But the uncertainty in the delay between real world events and
application programs reading the corresponding time is likely to be
rather larger.
Unless the interrupt path has a low latency, yes, but this is true of
any operating system. Choose OS according the the application needs, as
always. As the OP was talking minutes offset, there are other problems
to be solved first! I wonder whether a time-zone setting is incorrect?
(Also, at least the past, ntpd only compensated well for Windows on
lightly loaded systems.)
You can see how well or otherwise NTP works on my Windows systems here:
http://www.satsignal.eu/mrtg/performance_ntp.php#windows-stratum-1
I keep systems lightly loaded as far as possible. Perhaps PC Feenix is
the most heavily loaded of the systems there. Caveat, these are synced
to several local stratum-1 servers with a poll interval of 32 seconds,
as I wanted good timekeeping performance.
--
Cheers,
David
Web: http://www.satsignal.eu
_______________________________________________
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions
