[ntp:questions] improving ntpd performance on Windows

David J Taylor david-taylor at blueyonder.neither-this-bit.nor-this.co.uk
Thu Mar 5 07:47:01 UTC 2009


Dave Hart wrote:
> On Mar 2, 1:46 pm, "David J Taylor" <david-tay... at blueyonder.neither-
> this-bit.nor-this.co.uk> wrote:
>>
>> Folks,
>>
>> Dave Hart asked that I post this information:
>>
>> Very briefly, I happen to have a 100Hz system, and have been playing
>> with the improved version of ntpd.exe on a Windows 2000 system.
>> Working with Dave, we found that it did have some issues, and it
>> seems that these issues have probably been resolved by setting the
>> sample interval to 26 milliseconds instead of the default 43
>> milliseconds.
>
> After seeing the performance of David Taylor's Windows ntpd with a
> refclock on a 100 Hz system was not doing as well as my 64 Hz system,
> David spent a chunk of time testing all the intervals from 26 through
> 32 on his system and noting the jitter over a few minutes.  He found
> that 27 was the best NTPD_INT_INT value for a 100 Hz system of those
> tested.  David, would you mind posting your results in case anyone
> else wants to try more intervals and do so comparably?

Yes, Dave.  Here are the results:

_____________________________
Using the interactive start of ntpd.exe, I ran for a couple of minutes 
first (waiting for a 377 in the Reach field), then sampled the reported 
Jitter values from an "ntpq -p" at about half-minute intervals, with each 
INT value.

26 - jitter = 0.040 0.034 0.123 0.052 0.044

27 - jitter = 0.034 0.032 0.046 0.050 0.042

28 - jitter = 0.043 0.060 0.065 0.069 0.033

29 - jitter = 0.090 0.072 0.045 0.153 0.059

30 - jitter = 0.508 0.255 0.253 0.223 0.249

31 - jitter = 0.070 0.066 0.059 0.071 0.065

32 - jitter = 0.085 0.048 0.058 0.070 0.029

I wondered if the VNC remote control might be affecting the results, so I 
tried repeating the 27 with and without:

27 with VNC - jitter = 0.034 0.032 0.046 0.050 0.042

27 without VNC - jitter = 0.034 0.040 0.063 0.050 0.025

So it looks as if the VNC wasn't affecting the result.
_____________________________


This was with an old Windows 2000 Server PC, a 550Mhz Pentium III, so I 
wasn't expecting great results.  The current timekeeping can be seen here, 
and you can see that going from zero to near 100% CPU load has upset the 
timekeeping.

  http://www.satsignal.eu/mrtg/performance_bacchus.php

All my PCs are shown here:

  http://www.satsignal.eu/mrtg/performance_ntp.php

Cheers,
David 




More information about the questions mailing list