Re: [MEXT] GRE support in DSMIPv6

Hi Pasi, Hesham,

The TLV header was specified in the DS-MIPv6 document after rather a
long and acrimonious debate on the former MIP6 mailing list. There were
atleast two consensus calls that were run at that time. Anytime you have
a UDP header with IPv4/IPv6/GRE header following it, you need the TLV
header. At that time, there were folks arguing for using GRE
encapsulation with MIPv6 also. PMIPv6 IPv4 support was not the only
scenario for the TLV header. We are overturning that consensus now.
Maybe folks who were arguing for the TLV header with DS-MIPv6, are
either busy/not looked at this thread yet/or not on the MEXT mailing
list/etc.. :)

Moving the TLV header into a separate document at this point would
impact draft-ietf-netlmm-pmip6-ipv4-support. I don't think the TLV
header document can be standardized fast enough for
draft-ietf-netlmm-pmip6-ipv4-support to advance. One option would be to
move the TLV header and the text that describes how to negotiate it, to
either draft-ietf-netlmm-pmip6-ipv4-support or
draft-ietf-netlmm-grekey-option. 

My suggestion would be to leave the TLV header in the DS-MIPv6 document.
Have some text that says the following. If UDP encapsulation is used
with DS-MIPv6 port, there could be IPv4, IPv6, GRE or some other header
that might follow the UDP header. If there is anything other than the
IPv4 or IPv6 header, the TLV header would be required. The use of GRE or
some other protocol after the TLV header is not specified and is out of
scope in the DS-MIPv6 document.

Vijay 

> -----Original Message-----
> From: mext-bounces@ietf.org [mailto:mext-bounces@ietf.org] On 
> Behalf Of Pasi.Eronen@nokia.com
> Sent: Thursday, January 15, 2009 3:54 AM
> To: hesham@elevatemobile.com; mext@ietf.org
> Subject: Re: [MEXT] GRE support in DSMIPv6 - AD review
> 
> Hesham,
> 
> I would strongly suggest moving the whole TLV header text to the
> separate GRE document.
> 
> In particular, if you assign a number for GRE in this document,
> you either need to describe how it works here, or have a normative
> reference to the NETLMM spec.
> 
> Best regards,
> Pasi
> 
> > -----Original Message-----
> > From: ext Hesham Soliman [mailto:hesham@elevatemobile.com] 
> > Sent: 14 January, 2009 14:23
> > To: mext@ietf.org
> > Cc: Eronen Pasi (Nokia-NRC/Helsinki)
> > Subject: GRE support in DSMIPv6 - AD review
> > 
> > Folks, 
> > 
> > Part of Pasi's review for DSMIPv6 was a comment on the lack of
> > specification for GRE support in the spec. He said it was vastly
> > under-specified, no details on the tunnelling, setting of different
> > parts of the GRE header ...etc.
> > 
> > I suggested that we don't explicitly mention GRE in the spec but we
> > keep the TLV tunnelling format and reserve the numbers for NETLMM to
> > specify exactly how it will be used in a separate document. I think
> > you would agree that this is largely driven by NETLMM needs and we
> > shouldn't specify the details in MEXT. Pasi was ok with that.
> > 
> > Please express your opinion on this soon because Pasi's comments are
> > the last comments for the draft and I want to handle them by Monday
> > at the latest.
> > 
> > Please avoid discussing the merits of GRE....etc, the question is:
> > 
> > Are there any objections to removing explicit references to GRE
> > while reserving the numbers in the TLV header for it to be specified
> > clearly in NETLMM?
> > 
> > Thanks,
> > 
> > Hesham
> _______________________________________________
> MEXT mailing list
> MEXT@ietf.org
> https://www.ietf.org/mailman/listinfo/mext
> 
_______________________________________________
MEXT mailing list
MEXT@ietf.org
https://www.ietf.org/mailman/listinfo/mext