[ntp:questions] NTP absolute accuracy?

Todd Glassey tglassey at earthlink.net
Sat Oct 31 15:43:58 UTC 2009


David J Taylor wrote:
> "David J Taylor" 
> <david-taylor at blueyonder.not-this-bit.nor-this.co.uk.invalid> wrote in 
> message news:%XQGm.126$Ym4.58 at text.news.virginmedia.com...
>   
>> "Unruh" <> wrote in message news:34JGm.50898$PH1.36449 at edtnps82...
>> []
>>     
>>> Note that chrony will give you a factor of 2 or three improvement over
>>> ntp  in the errors, assuming that the roundtrip is equally split on
>>> Linux or BSD.
>>>       
True but it doesnt support AutoKEY operations so it has no value in 
provable time settings which are now key requirements in many commercial 
settings.
>> For those without wide-bandwidth academic connections - those folks on 
>> cable or ADSL - how good is an "equal split round trip" assumption?
>>
>> David
>>     
David -
This doesnt take into account the issues of peering agreements and 
backhauling which add a non-deterministic type of latency to the 
time-setting process as does the peering agreements between the various 
carriers who's networks the time-service providers systems operate in.  
It also takes into account system I/O latency's and the issues of 
time-provider or infrastructure loading.

Also most parties using cable networks are probably running home gear 
with SNTP enablement in most instances. Its unlikely that a NTP client 
exists in most sub-$100 infrastructure devices.

I think the answer is that no one knows and it could be as little as a 
few milliseconds or it could be more and we wont know. It also depends 
on whether the time server those client systems are tied to are located 
on the Open Internet or are operated by the Cable Providers for their 
clients use. Then the secondary non-deterministic impulse and the 
asymmetric routing delays are mitigated or eliminated so again - it 
depends on how NTP is being used and whether its a real NTP instance or 
just a SNTP instance (as it exists today) being used.
>
> Thanks, David, David and Jan.  A few milliseconds is what I had expected, 
> so if you are on a consumer line, what implications does that have for 
> unruh's comment?
>
> Cheers,
> David 
>
> _______________________________________________
> questions mailing list
> questions at lists.ntp.org
> https://lists.ntp.org/mailman/listinfo/questions
> ------------------------------------------------------------------------
>
>
> No virus found in this incoming message.
> Checked by AVG - www.avg.com 
> Version: 8.5.423 / Virus Database: 270.14.39/2470 - Release Date: 10/30/09 15:18:00
>
>   




More information about the questions mailing list