<?xml version="1.0" standalone="no"?>
                    <!DOCTYPE div SYSTEM "/www/backend/www-xml-443/dtd/caidaML.dtd">
                    <!-- do NOT ERASE the DOCTYPE declaration! --><div>


<tr bgcolor="#f4f4f4">
  <td>
<font face="helvetica,arial" size="2">
<b>URL:</b>
</font>
</td>
  <td>
<font face="helvetica,arial" size="2">
<a href="http://www.caida.org/publications/papers/2001/Rssac2001a/">http://www.caida.org/publications/papers/2001/Rssac2001a/</a>
</font>
  </td>
</tr>


<tr bgcolor="#e9e9e9">
  <td>
<font face="helvetica,arial" size="2">
<b>Entry Date:</b>
</font>
</td>
  <td>
<font face="helvetica,arial" size="2">
2003-10-02


</font>
  </td>
</tr>


<tr bgcolor="#f4f4f4">
  <td>
<font face="helvetica,arial" size="2">
<b>Abstract:</b>
</font>
</td>
  <td>
<font face="helvetica,arial" size="2">
We describe active measurements of topology and end-to-end latency
characteristics between several of the DNS root servers and a subset of
their clients using the skitter tool developed by CAIDA. We gather a sample
of clients for each monitored DNS root server, combine these samples into a
common target list and then actively probe these targets and analyze their
connectivity. We identify the subsets of destinations that have large
latency connections to all instrumented root name servers and discuss their
geographical make-up. Our goal is to build an analytical framework for
evaluating the optimality of root server placement and its impact on the
efficiency of the DNS service. The skitter tool and the methodology we
propose can also be used for monitoring the end-to-end performance in large
networks and for assessing the optimality of web servers placement in
general.



</font>
  </td>
</tr>
</div>

