Re: [trill] Tsvart early review of draft-ietf-trill-over-ip-10

Joe Touch <touch@isi.edu> Mon, 26 June 2017 18:42 UTC

Return-Path: <touch@isi.edu>
X-Original-To: trill@ietfa.amsl.com
Delivered-To: trill@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 5E70E129B17; Mon, 26 Jun 2017 11:42:54 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.901
X-Spam-Level:
X-Spam-Status: No, score=-6.901 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_HI=-5, RP_MATCHES_RCVD=-0.001] autolearn=ham autolearn_force=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 7MvGN6U9DXCD; Mon, 26 Jun 2017 11:42:52 -0700 (PDT)
Received: from vapor.isi.edu (vapor.isi.edu [128.9.64.64]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 3A49E1270AC; Mon, 26 Jun 2017 11:42:52 -0700 (PDT)
Received: from [128.9.184.87] ([128.9.184.87]) (authenticated bits=0) by vapor.isi.edu (8.13.8/8.13.8) with ESMTP id v5QIgIg6009301 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES128-SHA bits=128 verify=NOT); Mon, 26 Jun 2017 11:42:18 -0700 (PDT)
From: Joe Touch <touch@isi.edu>
To: Donald Eastlake <d3e3e3@gmail.com>, Magnus Westerlund <magnus.westerlund@ericsson.com>
Cc: tsv-art@ietf.org, "trill@ietf.org" <trill@ietf.org>, IETF Discussion <ietf@ietf.org>, draft-ietf-trill-over-ip.all@ietf.org
References: <149754795560.13109.17521244075940607817@ietfa.amsl.com> <CAF4+nEG-28weDot9R9Z4-05PX1tzBoKZSOHu8BJY2GiRzOv0nA@mail.gmail.com> <7cc18878-970a-efbe-8488-a4110b9b24e9@isi.edu>
Message-ID: <c006414f-71d9-6a4b-eaa2-c0e38e87e0ee@isi.edu>
Date: Mon, 26 Jun 2017 11:42:16 -0700
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:52.0) Gecko/20100101 Thunderbird/52.2.0
MIME-Version: 1.0
In-Reply-To: <7cc18878-970a-efbe-8488-a4110b9b24e9@isi.edu>
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
Content-Language: en-US
X-ISI-4-43-8-MailScanner: Found to be clean
X-MailScanner-From: touch@isi.edu
Archived-At: <https://mailarchive.ietf.org/arch/msg/trill/8JoZPIz3kxuwUmqj5EAFQxUry2Q>
Subject: Re: [trill] Tsvart early review of draft-ietf-trill-over-ip-10
X-BeenThere: trill@ietf.org
X-Mailman-Version: 2.1.22
Precedence: list
List-Id: "Developing a hybrid router/bridge." <trill.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/trill>, <mailto:trill-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/trill/>
List-Post: <mailto:trill@ietf.org>
List-Help: <mailto:trill-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/trill>, <mailto:trill-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 26 Jun 2017 18:42:54 -0000

PS - the idea that TCP segments within a single connection should ever
have different DSCPs is a good example of why it's a bad idea to even
'think' of TRILL over TCP as direct encapsulation. I.e., that concept is
inherently hazardous and should be avoided.

Joe


On 6/26/2017 10:15 AM, Joe Touch wrote:
> Hi, Donald,
>
>
> On 6/25/2017 5:07 PM, Donald Eastlake wrote:
>> Hi Magnus,
>>
>> Thanks for the extensive review. See my responses below.
>>
>> On Thu, Jun 15, 2017 at 1:32 PM, Magnus Westerlund
>> <magnus.westerlund@ericsson.com> wrote:
>>> Reviewer: Magnus Westerlund
>>> Review result: Not Ready
>>>
>>> Early review of draft-ietf-trill-over-ip-10
>>> Reviewer: Magnus Westerlund
>>> Review result: Not Ready
>>>
>>> TSV-ART review comments:
>>>
>>> I have set this to not ready as there are several issues, some significant that
>>> could affect the protocol realization significantly. Some may be me missing
>>> things in TRILL, I was not that familiar with it before this review and I have
>>> only tried looking up things, not reading the whole earlier specifications. So
>>> don't hesitate to push back and provide pointers to things that can resolve
>>> issues. The authors and the WG clearly have thought about a lot of issues and
>>> dealt with much already.
>> OK. Hopefully we can resolve these one way or the other.
>>
>> ...
>>> TCP Encapsulation issue
>>> -----------------------
>>>
>>> Section 5.6:
>>>
>>> The TCP encapsulation appear to be missing an delimiter format allowing each
>>> individual TRILL packet/payload to be read out of the TCP's byte stream. In
>>> other words, a normal implementation has no way of ensuring that the TCP
>>> payload starts with the start of a new TRILL payload. Multiple small TRILL
>>> payloads may be included in the same TCP payload, and also only parts as TCP is
>>> one way of dealing with TRILL packets that are larger than the IP+Encapsulation
>>> MTU that actually will work.
>>>
>>> This comment is based on that there appear to be no length fields included in
>>> the TRILL header. The most straight forward delimiter is a 2-byte length field
>>> for the TRILL payload to be encapsulated.
>> Right. It might also be useful to include some sort of check field, as
>> is done in BGP, to detect if you are out of sync in parsing the TCP
>> stream.
> There is nothing in BGP that ever assumes that TCP write boundaries are
> preserved. BGP uses markers and length fields to create message
> boundaries in TCP's bytestream. The same is needed here.
>
> Note that BGP also never claims to craft TCP packets by 'encapsulating'
> a BGP message in a TCP segment. That part of this document needs to be
> removed - it not how TCP is ever used.
>
>> Another point is that, while with UDP it seems fine to send packets
>> with assorted QoS, you don't want to encourage re-ordering of TCP
>> packets in a stream. So if TCP encapsulation is being used,
> Again - please, NO. NEVER use this term.
>
>> you want
>> to use the same DSCP value for the packets in a particular TCP stream.
> Again, this is nonsensical. TCP would set a DSCP for the connection,
> never in different ways for individual segments of a connection.
>
>
>> So, generally, you need to have a TCP connection per priority handling
>> category. Mapping the 8 priority levels into a smaller number of
>> handling categories is a normal thing to do so you certainly don't
>> necessarily need 8 TCP connections. Adding material on this should not
>> be too hard.
> Perhaps, but please - again, please - omit any mention or implication
> that this occurs via encapsulation.
>
> If you want to use TCP, please use it properly.
>
>>> Section 5.6:
>>>
>>> TCP endpoint requirements. I do wonder if an application like TRILL actual
>>> would need to discuss performance impacting implementation choices or
>>> limitations. For example use of NAGLE, the requirements on buffer sizes in
>>> relation to Bandwidth delay products, as buffer memory in a RBridge will impact
>>> performance.
>> Well, I'm not sure how deeply this document should get into such
>> performance issues. What about just saying something about
>> consideration being given to tuning TCP for performance and pointing
>> to one or a few other RFCs that talk about this?
> Because your use of TCP (even if changed to describe it correctly) isn't
> listed in those TCP RFCs.
>
> And it's not so simple - NAGLE helps performance for interactive systems
> that use single-byte messages (e.g., telnet) and reduces the number of
> outstanding "less than full" segments. When used for encapsulation,
> turning NAGLE off is the right thing for multibyte messages (e.g., TRILL
> messages) and can avoid the "gathering" delay (200 ms stalls when there
> isn't enough source data - i.e., incoming TRILL packets - to keep up
> with the outgoing segments), but could also generate a large number of
> small segments (which can interfere with segment-based congestion
> control, vs. ABC).
>
> Unless you want a very poorly performing result, *THIS* is what you need
> to drill down into.
>
> Joe
>
> _______________________________________________
> trill mailing list
> trill@ietf.org
> https://www.ietf.org/mailman/listinfo/trill