<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">On 29/07/12 12:11, Dobbins, Roland
wrote:<br>
</div>
<blockquote
cite="mid:CD5F7389-23E6-4C40-BA79-6AFCAC167878@arbor.net"
type="cite">...
<pre wrap="">Recursive & authoritative functionality should be separated, as well:
<a class="moz-txt-link-rfc2396E" href="http://dl.dropbox.com/u/25235895/dnstiers.jpg"><http://dl.dropbox.com/u/25235895/dnstiers.jpg></a>
</pre>
</blockquote>
<br>
Interesting diagram. I have a couple of questions:<br>
<ul>
<li>What is the significance of primary and secondary in that
diagram? I thought BIND 8+ had done away with all that, and
there were just masters and slaves. Is it a role only locally
significant on that site, like the anycast loopbacks?</li>
<li>What is the purpose of the external resolvers? Wouldn't their
purpose be just as easily fulfilled by the zone slaves?</li>
<li>What do the hidden masters and zone slaves use for external
resolution? The internal resolvers? One of the caching-only
groups?<br>
</li>
<li>What's the purpose of the aggregate caching-only forwarders?
Is it merely a scale issue that dictates their use instead of a
direct relationship between the caching-only forwarders and the
internal resolvers?<br>
</li>
</ul>
Sorry for the barrage of questions, but this is quite relevant to a
project i'm working on, and i'm keen to understand other people's
designs.<br>
<br>
Thanks in advance,<br>
Paul<br>
<br>
</body>
</html>