[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[tsvwg] draft-ietf-tsvwg-ecn-tunnel-04 - Comment on Section 7

I've just completed a review of the ECN Tunnel draft, and note that this is a much easier read than the last version and addresses my previous concerns. 

Best wishes,

Gorry

---

Section 7:
- I think the title doesn't draw the readers attention, IMHO "Future Non-Default Schemes" is too vague, and I think should explicitly mention "Alternate Semantics for ECN". 

- This section appears to clarify the brief text offered in RFC 4774.
- I interpret this section as offering some useful new guidance, but I'd urge care here. Currently it says: 

  "In one word the guidance is "Don't".  If a scheme requires tunnels to
   implement special processing of the ECN field for certain DSCPs, it
   is highly unlikely that every implementer of every tunnel will want
   to add the required exception and that operators will want to deploy
   the required configuration options.  Therefore it is highly likely
   that some tunnels within a network will not implement the required
   special case.  Therefore, designers of new protocols should avoid
   non-default tunnelling schemes if at all possible."
- I think to be explicit the guidance is "don't design alternative tunnel semantics" (I think you're not trying to make RFC4774 historic?)
- I also don't particularly like the "highly unlikely", an operator could use a single vendor solution that would be OK (perhaps) - to me, the core issue is the real concern that is hard to guarantee that all tunnel endpoints will have the same interpretation, and if I understand right, you are saying it is probable that some tunnels will not do what they should (which seems likely also to me).
- I think the section SHOULD conclude by saying this extends (clarifies, updates?) the few paras in RFC 4774 on tunnels, by concluding with something like: "This guidance extends the considerations in section 5.4 of RFC 4774, which requires new methods to consider the implications of tunnel encapsulated packets." ??? 
---