VRRPv3 IPv6 RFC-5798 checksum

joe mcfar <[email protected]> Mon, 7 May 2012 16:49:46 -0500
Newsgroups gmane.ietf.vrrp
Message-ID <CALsBhU9meNgbmh4vcXNHxeLNLxX6ER8-fQkbB-=Uu3fo2HGQyA@mail.gmail.com>
--===============2480495684684509843==
Content-Type: multipart/alternative; boundary=e0cb4efe2ee4ac106604bf793fb8

--e0cb4efe2ee4ac106604bf793fb8
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: quoted-printable

I am testing interoperability with one major vendor=92s VRRPv3 IPv6
implementation and was surprised to  find that the vendor=92s unit does not
include an IPv6 pseudo-header along with the VRRPv3 packet data when
calculating the checksum.   The current version of Wireshark (1.6.7) also
shows the VRRPv3 checksum as incorrect.  This would probably only be an
issue when trying to interoperate with another vendor=92s implementation.



I=92m interpreting RFC-5798 section 5.2.8 to say to calculate the checksum
over both the VRRPv3 packet data starting with the version field, AND over
the IPv6 pseudo-header (source, destination IPv6 addresses, upper layer
packet length, zero field, and next header of 112 for VRRP.)



Can someone on this mailing list verify my interpretation?



Thanks,

Joe McFarland

--e0cb4efe2ee4ac106604bf793fb8
Content-Type: text/html; charset=windows-1252
Content-Transfer-Encoding: quoted-printable



<p class=3D"MsoNormal">I am testing interoperability with one major vendor=
=92s VRRPv3
IPv6 implementation and was surprised to =A0find that the vendor=92s unit d=
oes
not include an IPv6 pseudo-header along with the VRRPv3 packet data when
calculating the checksum.=A0 =A0The current version of Wireshark (1.6.7)
also shows the VRRPv3 checksum as incorrect.=A0 This would probably only be
an issue when trying to interoperate with another vendor=92s implementation=
.</p>

<p class=3D"MsoNormal">=A0</p>

<p class=3D"MsoNormal">I=92m interpreting RFC-5798 section 5.2.8 to say to =
calculate
the checksum over both the VRRPv3 packet data starting with the version fie=
ld,
AND over the IPv6 pseudo-header (source, destination IPv6 addresses, upper
layer packet length, zero field, and next header of 112 for VRRP.) </p>

<p class=3D"MsoNormal">=A0</p>

<p class=3D"MsoNormal">Can someone on this mailing list verify my interpret=
ation?</p>

<p class=3D"MsoNormal">=A0</p>

<p class=3D"MsoNormal">Thanks,</p>

<p class=3D"MsoNormal">Joe McFarland</p>


--e0cb4efe2ee4ac106604bf793fb8--

--===============2480495684684509843==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
vrrp mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/vrrp

--===============2480495684684509843==--