Date: Fri, 5 Oct 2007 14:04:45 -0600 From: "Leino, Tammy" <tammy_leino@mentor.com> Subject: [dhcptest:00226] DHCPv6 1.0.7 self-test suite tests #46 and #50 broken To: <dhcptest@tahi.org> Message-Id: <2F83EB16CD718C43889CBCD31ECFD7F805AE83@na2-mail.mgc.mentorg.com> X-Mail-Count: 00226Hello All,In version 1.0.7 of the test suite for tests #46 and #50, the TAHI node =is unable to generate a proper RT when we reach the max RT for the =transaction. The value rolls over to zero instead of the proper value. =See the attached logs.Here is a snippit:12:43:18 got DHCPv6 Rebind MessageRebind MessageDHCPv6 Option Values Identifier Identifier = 8407889 Opt_DHCPv6_CID DUID-LL HardwareType = 1DUID-LL LinkLayerAddress= 00:00:11:22:33:44Opt_DHCPv6_ElapsedTime Time = 65535Opt_DHCPv6_IA_NA Option 1Identifier= 305419897T1= 0T2= 0#IA_Addr OptionAddr= 3ffe:501:ffff:100:200:11ff:fe22:3344PreferredLifetime= 2000ValidLifetime= 3000 RT = 300.76 12:49:47 got DHCPv6 Rebind MessageRebind MessageDHCPv6 Option Values Identifier Identifier = 8407889 Opt_DHCPv6_CID DUID-LL HardwareType = 1DUID-LL LinkLayerAddress= 00:00:11:22:33:44 Opt_DHCPv6_ElapsedTime Time = 65535 Opt_DHCPv6_IA_NA Option 1Identifier= 305419897T1= 0T2= 0#IA_Addr OptionAddr= 3ffe:501:ffff:100:200:11ff:fe22:3344PreferredLifetime= 2000ValidLifetime= 3000 13:00:17 Clear Captured Packets (Link0) RT = 0 The RT output of these tests worked properly in 1.0.5, although there =were other problems reported with the tests involving improper =assumptions of the next max RT. I think this test is still suffering =from improper calculations, since the RT of this last packet was over =the MRT, but this is allowable according to RFC 3315. When you reach =MRT, you reset RT to MRT + RAND * MRT. This could make the value up to =.10 * MRT greater than MRT.Best Regards,Tammy Leino226_2.html (attatchment)(tag is disabled)