Re: [tsvwg] Status of ECN encapsulation drafts (i.e., stuck)

[Bob Briscoe <[e-mail]>]

Jonathan,

On 13/03/2020 17:56, Jonathan Morton wrote:
>> On 13 Mar, 2020, at 7:45 pm, Bob Briscoe <[e-mail]> wrote:
>>
>> No, you have to read the email. The doubling comes at fragmentation, then the reassembly is meant to compensate for it.
> Ah, so the situation is that the CE mark occurs *before* the tunnel and the fragmentation.

No.

Are you reading my recent emails, or the main email?


Bob

>   But this should not cause any problem in the end, either, because the packets are reassembled before being interpreted by the transport receiver.  Reworking my earlier example:
>
> Let's consider two original, consecutive IP packets: A B, both marked ECT.  B is then marked CE by an AQM.  They are then encapsulated by a tunnel, producing TA(ECT) TB(CE).  These are both larger than the MTU, so are fragmented into TA1(ECT) TA2(ECT) TB1(CE) TB2(CE).  Eventually these packets are decapsulated, prior to which they must be reassembled, and the CE marks on TB1 and TB2 must be preserved somehow.
>
> TA1(ECT) TA2(ECT) always reassembles into TA(ECT), and TB1(CE) TB2(CE) reassembles into TB(CE) under RFC-3168 rules.  We would then have A(ECT) and B(CE) transmitted onward, which seems to match the intent of the AQM - in particular, the number of CE marks is preserved, not doubled.
>
> So again, what context am I missing?
>
>   - Jonathan Morton

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