IIRC, NTP actually sends a packet with a local timestamp, sends it to a
remote location for a remote timestamp, and then checks the local
timestamp when the packet is received to calculate RTT.<br>
<br>
-Richard<br>
<br>
<br>
<span type="cite">Steven M. Bellovin wrote:</span>
<p> </p>
<blockquote type="cite"
 style="border-left: thin solid blue; padding-left: 10px; margin-left: 0pt;"> <tt> <br>
In message
<a class="moz-txt-link-rfc2396E" href="mailto:Pine.LNX.4.44.0305140802170.7544-100000@twin.uoregon.edu">&lt;Pine.LNX.4.44.0305140802170.7544-100000@twin.uoregon.edu&gt;</a>, Joel
Jae <br>
ggli writes: <br>
&gt; <br>
&gt; <br>
&gt;Also if you just need&nbsp; a high level of syncronization between the
time on <br>
&gt;all your hosts you can just deploy one standalone ntp server, sync
it <br>
&gt;against public time sources and get everything synced against that.
its <br>
&gt;probably a 95% solution to most people's timeing needs. <br>
&gt; <br>
  <br>
If I recall correctly, NTP assumes that latency = RTT/2.&nbsp; You might <br>
make it work well for his application *if* you set up your tree so that<br>
your paths are each one hop, or at least symmetric over your network. <br>
  <br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; --Steve Bellovin, <a
 href="http://www.research.att.com/%7Esmb">http://www.research.att.com/~smb</a>
(me) <br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <a href="http://www.wilyhacker.com">http://www.wilyhacker.com</a>
(2nd edition of "Firewalls" book) <br>
  <br>
  <br>
  <br>
  </tt> </blockquote>