Index: [Article Count Order] [Thread]

Date: Wed, 9 Jul 2008 12:09:11 -0700
From: "Hunt Peter (Nokia-S&S/MtView)" <Peter.Hunt@nokia.com>
Subject: [users:00825] Question about test v6LC.2.2.7E in IPv6 Core suite 4-0-2
To: "users@tahi.org" <users@tahi.org>
Cc: "Hunt Peter (Nokia-S&S/MtView)" <Peter.Hunt@nokia.com>
Message-Id: <C87919B0-1576-4E7B-927C-DC11816FD38C@mimectl>
X-Mail-Count: 00825

  I'm running the Self Test suite in the following configuration:* Environment: FreeBSD 6.3* Test Tool version: v6eval 3.0.15* Self Test package version 4-0-2I have a question about the test v6LC.2.2.7E, which looks at the prefixes i=n Router Advertisements when a zero-ending prefix is configured.The test adds the prefix 8000::/64 to the NUT, and looks at the next RA gen=erated. In response, the NUT generates an RA with two prefixes: 3ffe:501:ff=ff:100::/64 and 8000::/64. The first prefix corresponds to the global addre=ss configured on the interface for all the tests.The test then fails because it fails to recognize the 3ffe:501:ffff:100::/6=4 prefix as valid.I can "fix" this problem by removing the 3ffe:501:ffff:100::/64 from the in=terface for the test, and re-configuring it again afterward. The NUT then g=enerates the RA with just the 8000::/64 prefix, and the test passes.However, the inclusion of both prefixes seems to be the correct behaviour. =From RFC 4861:      Prefix Information                     These options specify the prefixes that are on-link                     and/or are used for stateless address                     autoconfiguration.  A router SHOULD include all its                     on-link prefixes (except the link-local prefix) so                     that multihomed hosts have complete prefix                     information about on-link destinations for the                     links to which they attach.  If complete                     information is lacking, a host with multiple                     interfaces may not be able to choose the correct                     outgoing interface when sending traffic to its                     neighbors.Should the test be changed to allow the 3ffe:501:ffff:100::/64 in addition =to the 8000::/64 prefix? Or is expected that the NUT will remove the existi=ng global address when the 8000::/64 address is added? Or should the NUT no=t include both prefixes in the RA?Thanks in advance,Peter HuntSoftware EngineerNokia S&S
	

825_2.html (attatchment)(tag is disabled)