[ntp:questions] chrony and ntp comparison-- ADSL hookup

Unruh unruh-spam at physics.ubc.ca
Wed Feb 20 17:40:50 UTC 2008


"David J Taylor" <david-taylor at blueyonder.neither-this-bit.nor-this-bit.co.uk> writes:

>Unruh wrote:
>[]
>>> - the default for min and max poll are 6 and 10 (IIRC), and not 4
>>> and 7
>>
>>
>> And so? Exactly how is this supposed to make a difference?

>It means that your tests are not carried out on an "out of the box" 
>configuration, and therefore perhaps less useful than they might otherwise 
>be.  It's possible that some servers would see min/maxpoll of 4 and 7 as 
>abusive.

And no "out of box" test is possible. Neither ntp ( althoug it is much
closer) or chrony are set up to do so. 
Right now I do not care what servers find abusive. I am using my own
server, so I decide what is abusive. I am not advocating these parameters,
I am using them to test ntp and chrony. ALL of the tests indicate that
chrony disciplines the clock about a factor of 2-3 better than does ntp.
I have seen no evidence to the contrary, nor any reason to believe that any
different situation would turn that around. I think, but do not know, that
the habit of ntp of throwing away 85% of its measurements (clock-filter) is
behind this. I think that the tradeoff of reducing the effect of delay spikes
 on the timing is more than outweighed by using fewer data points to beat
down the statistics of the noise. That combined with the long time constant
of ntp which makes it badly behaved in the case of drift fluctuations is
what I believe makes ntp worse than chrony.
Chrony is by no means perfect. It is complicated, at present it works only
under Linux essentially, it has some design weirnesses as well. But I was
interested to test the very different design philosophies of chrony against
ntp. (Recall that chrony was written by one person, and was more or less
completed about 10 years ago and has not had much if any tuning since. It
is a pretty impressive piece of work).



>Testing with the ADSL hookup is a useful extra, though.

>Cheers,
>David 





More information about the questions mailing list