Re: [tsvwg] [Int-area] New draft to update L2TP, GRE, PPTP, GTP & VXLAN, etc. for ECN

Bob,

On Jul 8, 2016, at 5:41 PM, Bob Briscoe <ietf@bobbriscoe.net<mailto:ietf@bobbriscoe.net>> wrote:

Carlos,

On 08/07/16 20:32, Carlos Pignataro (cpignata) wrote:
Bob,

When you say L2TP, do you mean only L2TPv2 [RFC 2661] as in the I-D, or also L2TPv3 [RFC 3931]? (I think should be both).
Thx. I've added L2TPv3 to my local copy - might post -01 later tonight. I recall adding that to my list in the past, but it must have fallen off again.

One additional point of clarification — the draft says:

   This specification therefore updates the
   following specifications of tightly coupled shim headers by adding
   that RFC 6040 SHOULD apply when the shim header is used between IP
   headers:

However, some of the listed tunneling technologies include additional encapsulations between the shim and the inner IP. Those are not part of the shim header per se. For example, IP | GRE | Ethernet | IP. Same for VXLAN. There’s PPP for L2TP, etc. The draft scope says:

   In many cases the shim header(s) and the outer IP header are always
   added (or removed) as part of the same process.  We call this a
   tightly coupled shim header.

It might be beneficial to tighten the scope to more definitively spec if it is IP | shim | something | IP, for “something” being non-null.
Any suggestions for how I can make the scope clearer - I thought I had made it clear: it's only if the shim and outer IP are added to an inner IP. Because, in the example you give an Ethernet header doesn't have an ECN field.

That case falls under the last catch-all paragraph that refers to draft-ietf-ecn-encap-guidelines, which attempts to cover every possibility in a more general way. In section 4.2 & section 6 you'll find that if ECN is in an IP header within an Ethernet header either you give up, or if you're a switch-router you violate layering and look inside the Ethernet header to find ECN in the IP header. Then, you refer to RFC6040 to propagate to/from the outer IP, jumping over the shim.

Thanks. Indeed, the paragraphs in ietf-tsvwg-ecn-encap-guidelines cover my question.

Thanks,

— Carlos.

Bob

However,

Thanks,

— Carlos.

On Jul 8, 2016, at 8:18 AM, Bob Briscoe <ietf@bobbriscoe.net<mailto:ietf@bobbriscoe.net>> wrote:

tsvwg, intarea, and respective co-chairs,
[re-sent with hyphen in int-area@]

I have posted a new very brief draft (under 2 pages not incl. boilerplate), intended for standards track as a bis to RFC6040.
As suggested in Buenos Aires, this has been extracted from draft-ietf-tsvwg-ecn-encap-guidelines, to cut all the clutter and highlight solely the standards track stuff.

Propagating Explicit Congestion Notification Across IP Tunnel Headers Separated by a Shim
https://tools.ietf.org/html/draft-briscoe-tsvwg-rfc6040bis-00

If approved, it is intended to update L2TP, GRE, PPTP, GTP & VXLAN.
Obviously the IETF does not control GTP - see text for how 3GPP might use this spec.
Simialrly, given VXLAN is informational, it's perhaps not appropriate to update it - exactly how this is worded is for discussion.

For IETF-96, I've asked for a slot in intarea to explain, and I also hope to cover this in an ecn-encap-guidelines slot in tsvwg.

I'd appreciate help identifying more tunnelling protocols that follow the pattern of a shim sandwiched between two IP headers.
Since posting, I've looked at Joe's Tunnelling draft and realised I missed out Geneve and GUE. More?

Cheers

Bob

On 08/07/16 12:41, internet-drafts@ietf.org wrote:
A new version of I-D, draft-briscoe-tsvwg-rfc6040bis-00.txt
has been successfully submitted by Bob Briscoe and posted to the
IETF repository.

Name: draft-briscoe-tsvwg-rfc6040bis
Revision: 00
Title: Propagating Explicit Congestion Notification Across IP Tunnel Headers Separated by a Shim
Document date: 2016-07-08
Group: Individual Submission
Pages: 5
URL:            https://www.ietf.org/internet-drafts/draft-briscoe-tsvwg-rfc6040bis-00.txt
Status:         https://datatracker.ietf.org/doc/draft-briscoe-tsvwg-rfc6040bis/
Htmlized:       https://tools.ietf.org/html/draft-briscoe-tsvwg-rfc6040bis-00

Abstract:
   RFC 6040 on "Tunnelling of Explicit Congestion Notification" made the
   rules for propagation of ECN consistent for all forms of IP in IP
   tunnel.  This specification extends the scope of RFC 6040 to include
   tunnels where two IP headers are separated by a shim header that
   cannot stand alone.

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

--
________________________________________________________________
Bob Briscoe                               http://bobbriscoe.net/

_______________________________________________
Int-area mailing list
Int-area@ietf.org
https://www.ietf.org/mailman/listinfo/int-area

--
________________________________________________________________
Bob Briscoe                               http://bobbriscoe.net/