[ntp:questions] Re: Fingerprinting hosts by clock skew

Brian Inglis Brian.Inglis at SystematicSW.Invalid
Thu Mar 10 15:14:13 UTC 2005


On Thu, 10 Mar 2005 10:41:56 +0100 in comp.protocols.time.ntp, Brad
Knowles <brad at stop.mail-abuse.org> wrote:

>At 10:34 AM +0100 2005-03-10, Mxsmanic wrote:
>
>>  But if you applied the corrections directly to the clock hardware, then
>>  anything using the hardware clock would incorporate those corrections,
>>  including TCP/IP clocks ... right?
>
>	I don't think so.  The TCP stack on most OSes seems to maintain a 
>totally separate clock that is not influenced by the system clock, 
>the hardware clock, or anything else outside of the TCP stack.  This 
>is the disconnect that I believe needs to be corrected.

Possibly because they need a clock that is not affected by adjtime for
the protocol to work? 

-- 
Thanks. Take care, Brian Inglis 	Calgary, Alberta, Canada

Brian.Inglis at CSi.com 	(Brian[dot]Inglis{at}SystematicSW[dot]ab[dot]ca)
    fake address		use address above to reply



More information about the questions mailing list