Hi all, First, I'm glad to see that more and more people have realized the need for a protocol type field in the VXLAN header since I mentioned that need one year ago (http://list-archives.org/2012/09/21/nvo3-ietf-org/unified-encapsulation-headers-for-both-layer2-and-layer3-vpn-overlays/f/6013985816).
Second, I guess most people who have realized the above need would agree that the protocol type field would be a necessary field (like the Protocol Type field in the GRE header), rather than an optional field (like the Checksum field in the GRE header) in the long run. Both the usage of P bit as described in your draft and the interpretation of 0x0000 in the Protocol Type field as Ethernet payload as described in our draft are just intended for backward compatibility to the early implementation of the pre-standard VXLAN. In order to reduce the potential risk of compatibility in the future, I feel it'd better for the NVo3 WG to make an early decision upon such extension to VXLAN before the WG adoption of the existing VXLAN data format. Best regards, Xiaohu > -----邮件原件----- > 发件人: [email protected] [mailto:[email protected]] 代表 Paul > Quinn (paulq) > 发送时间: 2013年9月10日 10:03 > 收件人: Lucy yong > 抄送: [email protected] > 主题: Re: [nvo3] New Version Notification for draft-quinn-vxlan-gpe-00.txt > > Lucy, > > Thank you for your comments. > > The P bit provides much needed flexibility for the protocol. First and > foremost, > it enables the creation of a single on the wire protocol that encompasses both > VXLAN and LISP (please see http://tools.ietf.org/html/draft-lewis-lisp-gpe-00 > for > the associated LISP draft). The P bit also enables the evolution of the > protocol, > allowing the format to change as needed with a simple bit change. > > Thank you, > Paul > > > On Sep 6, 2013, at 3:20 PM, Lucy yong <[email protected]> wrote: > > > Hi Paul, > > > > I read this draft. Both this draft and our draft > http://datatracker.ietf.org/doc/draft-yong-l3vpn-nvgre-vxlan-encap/ propose > adding prototype type field in VXLAN header in supporting L3 overlays (or > other) > beside L2 overlays. The only difference between two is that this draft > requests > using P bit to indicate the presence of protocol type field in VXLAN header. > We > don't think this is necessary. For the backward compatibility, we propose > treating protocol type value 0 as Ethernet payload. > > > > This draft gives two ways to do L2 overlay. If P bit is set and protocol > > type > value 0x6558, L2 payload is follow, or P bit is unset, L2 payload is follow. > IMHO, > this is bad. > > > > Regards, > > Lucy > > > > > > > > -----Original Message----- > > From: [email protected] [mailto:[email protected]] On Behalf Of > Paul Quinn (paulq) > > Sent: Friday, September 06, 2013 5:04 PM > > To: [email protected] > > Subject: [nvo3] Fwd: New Version Notification for > > draft-quinn-vxlan-gpe-00.txt > > > > FYI for the group. > > > > Paul > > > > Begin forwarded message: > >> > >> A new version of I-D, draft-quinn-vxlan-gpe-00.txt has been > >> successfully submitted by Paul Quinn and posted to the IETF > >> repository. > >> > >> Filename: draft-quinn-vxlan-gpe > >> Revision: 00 > >> Title: Generic Protocol Extension for VXLAN > >> Creation date: 2013-08-27 > >> Group: Individual Submission > >> Number of pages: 13 > >> URL: > http://www.ietf.org/internet-drafts/draft-quinn-vxlan-gpe-00.txt > >> Status: http://datatracker.ietf.org/doc/draft-quinn-vxlan-gpe > >> Htmlized: http://tools.ietf.org/html/draft-quinn-vxlan-gpe-00 > >> > >> > >> Abstract: > >> This draft describes a mechanism for adding multi-protocol support to > >> Virtual eXtensible Local Area Network (VXLAN). Protocol > >> identification is carried in the VXLAN header and is used to describe > >> the encapsulated payload. > >> > >> > >> > >> > >> Please note that it may take a couple of minutes from the time of > >> submission until the htmlized version and diff are available at > >> tools.ietf.org. > >> > >> The IETF Secretariat > >> > > > > _______________________________________________ > > nvo3 mailing list > > [email protected] > > https://www.ietf.org/mailman/listinfo/nvo3 > > _______________________________________________ > nvo3 mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/nvo3 _______________________________________________ nvo3 mailing list [email protected] https://www.ietf.org/mailman/listinfo/nvo3
