Re: [nvo3] [Int-area] Fwd: New Version Notification for draft-ietf-nvo3-gue-03.txt

On Thu, Jun 16, 2016 at 6:48 PM, Xuxiaohu <xuxiaohu@huawei.com> wrote:
>
>
>> -----Original Message-----
>> From: Tom Herbert [mailto:tom@herbertland.com]
>> Sent: Thursday, June 16, 2016 10:37 PM
>> To: Xuxiaohu
>> Cc: int-area@ietf.org; nvo3@ietf.org
>> Subject: Re: [nvo3] [Int-area] Fwd: New Version Notification for
>> draft-ietf-nvo3-gue-03.txt
>>
>> On Thu, Jun 16, 2016 at 2:12 AM, Xuxiaohu <xuxiaohu@huawei.com> wrote:
>> >
>> >
>> >> -----Original Message-----
>> >> From: Int-area [mailto:int-area-bounces@ietf.org] On Behalf Of Tom
>> >> Herbert
>> >> Sent: Saturday, June 11, 2016 1:21 AM
>> >> To: int-area@ietf.org; nvo3@ietf.org
>> >> Subject: [Int-area] Fwd: New Version Notification for
>> >> draft-ietf-nvo3-gue-03.txt
>> >>
>> >> Hello,
>> >>
>> >> We've posted a new version of GUE. The primary addition is that we
>> >> added GUE version 1 that allows direct encapsulation of IPv4 and IPv6
>> >> over UDP (port 6080). This effectively implements
>> >> draft-xu-intarea-ip-in-udp-03 also.
>> > Tom,
>> >
>> > What's the real benefit of such implementation of IP-in-UDP compared to the
>> approach as described draft-xu-intarea-ip-in-udp-03? Save one UDP port
>> number?
>> >
>> Yes, saves a port number.
>
> To save a port number, the header format is made ugly. Is it worthwhile? If UDP port resource was so sparse as you had imagined, I think the UDP port resource keeper would not allocate two different port numbers for VXLAN and VXLAN-GPE since the P-bit in VXLAN-GPE header is enough to distinguish VXLAN-GPE from VXLAN. For more details, please look at section 3.2 of (https://tools.ietf.org/html/draft-ietf-nvo3-vxlan-gpe-02#page-6).
>
Xiaohu,

Ugliness is in the eye of the beholder. Personally, I think this
solution is clever and elegant. Also, this was a trivial code change
to make work in GUE.

As for the UDP port number space, I suggest you review RFC6335. From that:

"Conservation of the port number space is required because this space
is a limited resource, so applications are expected to participate in
the traffic demultiplexing process where feasible.  The port numbers
are expected to encode as little information as possible that will
still enable an application to perform further demultiplexing by
itself."

To flip the question around, what would be the value be for allocating
new port numbers to do IP-over-UDP if this can already be done with an
existing port number? The fact that this same port number can be also
used for GUE isn't particularly relevant (i.e. the Swiss Army Knife
point), an implementation is perfectly free to only send IP-in-UDP on
the port and never has to send a single GUE packet (vers. 0).
Semantically and operationally there is no difference between port
6080 or using a new port for IP-in-UDP.

Tom

>> >
>> >> Thanks,
>> >> Tom
>> >>
>> >> ---------- Forwarded message ----------
>> >> From:  <internet-drafts@ietf.org>
>> >> Date: Fri, Jun 10, 2016 at 10:14 AM
>> >> Subject: New Version Notification for draft-ietf-nvo3-gue-03.txt
>> >> To: Tom Herbert <tom@herbertland.com>, Lucy Yong
>> >> <lucy.yong@huawei.com>, Osama Zia <osamaz@microsoft.com>
>> >>
>> >>
>> >>
>> >> A new version of I-D, draft-ietf-nvo3-gue-03.txt has been
>> >> successfully submitted by Tom Herbert and posted to the IETF repository.
>> >>
>> >> Name:           draft-ietf-nvo3-gue
>> >> Revision:       03
>> >> Title:          Generic UDP Encapsulation
>> >> Document date:  2016-06-10
>> >> Group:          nvo3
>> >> Pages:          28
>> >> URL:
>> >> https://www.ietf.org/internet-drafts/draft-ietf-nvo3-gue-03.txt
>> >> Status:         https://datatracker.ietf.org/doc/draft-ietf-nvo3-gue/
>> >> Htmlized:       https://tools.ietf.org/html/draft-ietf-nvo3-gue-03
>> >> Diff:           https://www.ietf.org/rfcdiff?url2=draft-ietf-nvo3-gue-03
>> >>
>> >> Abstract:
>> >>    This specification describes Generic UDP Encapsulation (GUE), which
>> >>    is a scheme for using UDP to encapsulate packets of arbitrary IP
>> >>    protocols for transport across layer 3 networks. By encapsulating
>> >>    packets in UDP, specialized capabilities in networking hardware for
>> >>    efficient handling of UDP packets can be leveraged. GUE specifies
>> >>    basic encapsulation methods upon which higher level constructs, such
>> >>    tunnels and overlay networks for network virtualization, can be
>> >>    constructed. GUE is extensible by allowing optional data fields as
>> >>    part of the encapsulation, and is generic in that it can encapsulate
>> >>    packets of various IP protocols.
>> >>
>> >>
>> >>
>> >>
>> >> 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
>> >>
>> >> _______________________________________________
>> >> Int-area mailing list
>> >> Int-area@ietf.org
>> >> https://www.ietf.org/mailman/listinfo/int-area
>> >
>> > _______________________________________________
>> > nvo3 mailing list
>> > nvo3@ietf.org
>> > https://www.ietf.org/mailman/listinfo/nvo3