Index: [Article Count Order] [Thread]

Date: Fri, 14 Dec 2007 07:10:04 -0700
From: "Leino, Tammy" <tammy_leino@mentor.com>
Subject: [users:00445] Re: Release 1.5.0b1 Phase-2 PMTU - Test #9 Host Test
To: "Yukiyo Akisada" <akisada@tahi.org>
Cc: <users@tahi.org>
Message-Id: <2F83EB16CD718C43889CBCD31ECFD7F81B8801@na2-mail.mgc.mentorg.com>
In-Reply-To: <20071214104229.fbb84a47.akisada@tahi.org>
X-Mail-Count: 00445

Thank you Yukiyo.

I have read this section many times trying to make it fit into test #9,
but I still cannot make sense of it.  It doesn't make sense that the
extra 8 bytes is a fragment header, because I cannot see why the packet
needs to be fragmented.  Can you please explain why the packet needs to
be fragmented.

Best Regards,
Tammy Leino


-----Original Message-----
From: Yukiyo Akisada [mailto:akisada@tahi.org]
Sent: Thursday, December 13, 2007 8:42 PM
To: Leino, Tammy
Cc: users@tahi.org
Subject: Re: [users:00436] Re: Release 1.5.0b1 Phase-2 PMTU - Test #9
Host Test


Hi, Tammy.

Please check RFC 1981, Chapter 4.

    241       Note: A node may receive a Packet Too Big message
reporting a
    242       next-hop MTU that is less than the IPv6 minimum link MTU.
In that
    243       case, the node is not required to reduce the size of
subsequent
    244       packets sent on the path to less than the IPv6 minimun
link MTU,
    245       but rather must include a Fragment header in those packets
[IPv6-
    246       SPEC].

In this case, extra 8 bytes mean Fragment header.

And actual test specification is
<http://www.ipv6ready.org/announcement/public_review20071204_p2core.html
>.

Perldoc isn't the master.
We are trying to follow the update of master specification,
but sometimes perldoc contains old information.

Please use the specification located at ipv6ready.org.

Regards,


On Wed, 12 Dec 2007 07:18:23 -0700
"Leino, Tammy" <tammy_leino@mentor.com> wrote:

> Hello All,
>  
> I reported this issue last Thursday and have not yet heard back from
anyone.  Do you require further clarification on any points?  If so,
please let me know.
>  
> Best Regards,
> Tammy Leino
>  
> 
> -----Original Message-----
> From: Leino, Tammy 
> Sent: Thursday, December 06, 2007 4:42 PM
> To: users@tahi.org
> Subject: [users:00426] Release 1.5.0b1 Phase-2 PMTU - Test #9 Host
Test
> 
> 
> 
> Hello All, 
> 
> There is a discrepancy in the test procedure diagram and description
for this test.  The MTU in the Packet Too Big message is actually 56 for
this test, and not zero as shown in the procedure diagram.
> 
>    TN2                    TR1                    NUT
>     |                      |                      |
>     |-------------------------------------------->|
>     |   1.Echo Request     |                      |
>     |   (1280 octets)      |                      |
>     |                      |                      |
>     |<--------------------------------------------|
>     |   2.Echo Reply       |                      |
>     |   (1280 octets)      |                      |
>     |                      |                      |
>     |                      +--------------------->|
>     |                      |  3.Packet Too Big    |
>     |                      |    (MTU 0)           |
>     |                      |                      |
>     |-------------------------------------------->|
>     |   4.Echo Request     |                      |
>     |   (1280 octets)      |                      |
>     |                      |                      |
>     |<--------------------------------------------|
>     | 5.Echo Reply         |                      |
>     |   (1288 octets)      |                      |
>     |                      |                      |
>     v                      v                      v 
> 
> Also, can someone please explain to me why the NUT would echo 1288
bytes when it receives only 1280 bytes?  I am referring to #4 and #5 of
the above diagram.
> 
> I am having trouble understanding why this test expects the NUT to
fragment the echo reply when the packet size is equal to the IPv6
minimum link MTU.  We do not fragment until we reach the link MTU of
1500.  Even if we set the link MTU to 1280, the IPv6 minimum link MTU,
we should not fragment a 1280 byte echo reply packet.  I'm sure I am
overlooking something here.  Your advice is greatly appreciated.
> 
> Best Regards,
> Tammy Leino 
> 
> 


------------------------------------------------------------------------
Yukiyo Akisada <akisada@tahi.org>