]> git.ipfire.org Git - thirdparty/bind9.git/commitdiff
more changes from Jeremy
authorMark Andrews <marka@isc.org>
Tue, 29 May 2007 23:12:42 +0000 (23:12 +0000)
committerMark Andrews <marka@isc.org>
Tue, 29 May 2007 23:12:42 +0000 (23:12 +0000)
doc/arm/Bv9ARM-book.xml

index 95f89b726ee00b77478a732907da1231efc6f547..e8223c284831226ac73350b5d05b9cdde1f67b21 100644 (file)
@@ -18,7 +18,7 @@
  - PERFORMANCE OF THIS SOFTWARE.
 -->
 
-<!-- File: $Id: Bv9ARM-book.xml,v 1.326 2007/05/28 00:09:07 marka Exp $ -->
+<!-- File: $Id: Bv9ARM-book.xml,v 1.327 2007/05/29 23:12:42 marka Exp $ -->
 <book xmlns:xi="http://www.w3.org/2001/XInclude">
   <title>BIND 9 Administrator Reference Manual</title>
 
         traffic.
         Additionally, if additional section caching
         (<xref linkend="acache"/>) is enabled,
-        the <command>max-acache-size</command> can be used to
+        the <command>max-acache-size</command> option can be used to
         limit the amount
         of memory used by the mechanism.
         It is still good practice to have enough memory to load
@@ -4283,26 +4283,27 @@ category notify { null; };
                      DNS due to timeouts.  This is often due to
                      the remote servers not being RFC 1034 compliant
                      (not always returning FORMERR or similar to
-                     EDNS queries and other extension to the DNS
-                     when they are not understood).  i.e. this is
+                     EDNS queries and other extensions to the DNS
+                     when they are not understood).  In other words, this is
                      targeted at servers that fail to respond to
                      DNS queries that they don't understand.
                    </para>
                    <para>
-                     Note: the log message can be also due to
+                     Note: the log message can also be due to
                      packet loss.  Before reporting servers for
-                     non RFC 1034 compliance they should be re-tested
-                     to determine the nature of of the non-compliance.
-                     This testing should prevent / reduce the
-                     number of false positive reports.
+                     non-RFC 1034 compliance they should be re-tested
+                     to determine the nature of the non-compliance.
+                     This testing should prevent or reduce the
+                     number of false-positive reports.
                    </para>
                    <para>
                      Note: eventually named will have to stop
                      treating such timeouts as due to RFC 1034 non
                      compliance and start treating it as plain
-                     packet loss as falsely classifying packet
+                     packet loss.  Falsely classifying packet
                      loss as due to RFC 1034 non compliance impacts
-                     on DNSSEC validation.
+                     on DNSSEC validation which requires EDNS for
+                     the DNSSEC records to be returned.
                    </para>
                  </entry>
                </row>