Re: [multipathtcp] AD review is coming

[<[e-mail]>]

About 
>    Thus, the listener will ignore the MP_CAPABLE of this SYN-segment and
>    reply with a SYN/ACK that does not include an MP_CAPABLE, thus
>    leading to a fallback to regular TCP.  An initiator MAY cache this
>    information about a peer and for future connections, MAY choose to
>    attempt using MPTCP v0, if supported, before recording the host as
>    not supporting MPTCP.

I think one problem is the phrase "thus leading to a fallback to regular TCP", which is a bit misleading. Suggest deleting and adding more explanation on these lines:

Thus, the listener will ignore the MP_CAPABLE of this SYN-segment and reply with a SYN/ACK that does not include an MP_CAPABLE. The initiator MAY choose to immediately fall back to TCP or MAY choose to attempt a connection using MPTCP v0 (if the initiator supports v0), in order to discover whether the listener supports the earlier version of MPTCP. In general a MPTCP v0 connection is preferred to a TCP one, however in a particular deployment scenario it may be known that the listener is unlikely to support MPTCPv0 and so the initiator may prefer not to attempt a v0 connection. 
An initiator MAY cache information for a peer about what version of MPTCP it supports if any, and use this information for future connection attempts.

phil

-----Original Message-----
From: Mirja Kuehlewind [mailto:[e-mail]] 
Sent: 16 April 2019 12:17
To: Alan Ford <[e-mail]>
Cc: [e-mail]; [e-mail]
Subject: Re: AD review is coming

Hi Alan,

Yes, I was looking also for eventually a tiny update to the doc (while it would be most important to mention this in the write-up). I think the one thing that could be added below, is one more sentence or maybe even halve a sentence saying that this fallback to plain TCP (instead of v0) is okay, as general deployment is low and there is also an expectation that server get updated first or entities using MPTCP today actually have some control to update both client and server. Or whatever you think is right to say. So the question to answers is why is it okay that fallback is to plain TCP instead of v0…?

Mirja


> On 15. Apr 2019, at 18:55, Alan Ford <[e-mail]> wrote:
> 
> Hi Mirja,
> 
> To be clear, you are looking for some rationale in the draft text as well? The technical discussion of the fallback process is given in Section 3.1:
>    The MP_CAPABLE exchange in this specification (v1) is different to
>    that specified in v0 [RFC6824].  If a host supports multiple versions
>    of MPTCP, the sender of the MP_CAPABLE option SHOULD signal the
>    highest version number it supports.  In return, in its MP_CAPABLE
>    option, the receiver will signal the version number it wishes to use,
>    which MUST be equal to or lower than the version number indicated in
>    the initial MP_CAPABLE.  There is a caveat though with respect to
>    this version negotiation with old listeners that only support v0.  A
>    listener that supports v0 expects that the MP_CAPABLE option in the
>    SYN-segment includes the initiator's key.  If the initiator however
>    already upgraded to v1, it won't include the key in the SYN-segment.
>    Thus, the listener will ignore the MP_CAPABLE of this SYN-segment and
>    reply with a SYN/ACK that does not include an MP_CAPABLE, thus
>    leading to a fallback to regular TCP.  An initiator MAY cache this
>    information about a peer and for future connections, MAY choose to
>    attempt using MPTCP v0, if supported, before recording the host as
>    not supporting MPTCP.
> 
> But you would like to see some discussion mentioning the migration from Experimental to Proposed Standard? And the benefits it brings? Would that cover it?
> 
> Many thanks,
> Alan
> 
>> On 12 Apr 2019, at 17:40, Mirja Kuehlewind <[e-mail]> wrote:
>> 
>> Hi authors, hi shepherd/Phil,
>> 
>> Just wanted to quickly let you know that I’ve started the IETF last call just now in order to hopefully get this document on the telechat in 3 weeks. 
>> 
>> I’ve not finished my AD review completely but I am nearly done and convinced that this doc is ready to start IETF last call. So far I have only some editorial or smallish comments and nits which I will send in detail beginning of next week and can be address during or after IETF last call.
>> 
>> Only one request for now (mostly with an eye on the IESG evaluation): The draft says that changes are made which are not backward compatible (see sec 3.1). That’s fine, however, it would be good to also explain a little bit in the draft as well as in the shepherd write up (!) why this is okay. My understanding is that in case of a v0 server the connection will “just" fall-back to non-MPTCP capable and that this is acceptable in current deployment situation. Please confirm and at least update the write-up accordingly as soon as possible!
>> 
>> Thanks!
>> Mirja
>> 
>> 
>> 
>