Re: [TICTOC] Enterprise profile update

Wojciech Owczarek <wojciech@owczarek.co.uk> Tue, 08 April 2014 22:19 UTC

Return-Path: <wojciech@owczarek.co.uk>
X-Original-To: tictoc@ietfa.amsl.com
Delivered-To: tictoc@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id B55B41A0752 for <tictoc@ietfa.amsl.com>; Tue, 8 Apr 2014 15:19:10 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.078
X-Spam-Level:
X-Spam-Status: No, score=-0.078 tagged_above=-999 required=5 tests=[BAYES_40=-0.001, FM_FORGED_GMAIL=0.622, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_LOW=-0.7] autolearn=ham
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 YIxCRkShERjO for <tictoc@ietfa.amsl.com>; Tue, 8 Apr 2014 15:19:06 -0700 (PDT)
Received: from mail-vc0-f171.google.com (mail-vc0-f171.google.com [209.85.220.171]) by ietfa.amsl.com (Postfix) with ESMTP id CF8701A0756 for <tictoc@ietf.org>; Tue, 8 Apr 2014 15:19:03 -0700 (PDT)
Received: by mail-vc0-f171.google.com with SMTP id lg15so1407032vcb.30 for <tictoc@ietf.org>; Tue, 08 Apr 2014 15:19:03 -0700 (PDT)
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=+yWU1i8lK89VDG/SfeUJ3W6Zd6j8uIVxDAecNmEczf8=; b=ekuZ9UBvS8hOs9WjfhVIivGEVMoz0HXle+SVzN/3pxCi5qGmXCaCHxvM6CZzy+5iRV WaxH65huZfLF4sH6wvTJJy/etjVWs+FT7rTpFzJ3DQvbeU/B+h6FsWdFG6h4hPA6h9hT civrSMQKHAa8NGOuKpu89vnySXV0+XO3ipFxIPmaHlFnkYKCkhpuDmJmru6aBQ5eMdyP hfQkvoGCxHypfuNMS5ozSt97uybrh5wDkF1U74TdHI22hCzb2tqry7ZUUXQQfjQDzHVh d9DFbj9nu3aIJD8QlMg7hSEmzu5pNGXdAm0Q4xUIaCNnlkdVLX8941F4C7GuIwUWnqlY BWdw==
X-Gm-Message-State: ALoCoQlut9xxrcF5koP/Hwy4FkcJFXL11rBqCzKuiyQ/8vTYguIUayr/c7kmBKt2ViXXiVxPIrWg
MIME-Version: 1.0
X-Received: by 10.221.28.202 with SMTP id rv10mr5408051vcb.10.1396995542261; Tue, 08 Apr 2014 15:19:02 -0700 (PDT)
Received: by 10.58.46.44 with HTTP; Tue, 8 Apr 2014 15:19:02 -0700 (PDT)
In-Reply-To: <CALw1_Q1ObS652A+GazmmB4gfkNnnJ4DcsiJ5PoQCd6wxNw-mrA@mail.gmail.com>
References: <CACQYgzE106JaKArgc=3HKkKcrdvSy5PmK-A0=_ZMdfrFUJbrtQ@mail.gmail.com> <20140405112034.GM22106@netboy> <20140405133748.GB4566@netboy> <CALw1_Q3hNRNpHzQ=1z0XRTqvcvm2W7=Gm_ZWa2UJT-BqYTvmZw@mail.gmail.com> <CF68E7A8.7488B%lmontini@cisco.com> <1D3F6A7C-AD27-4A63-9F63-41D9929AE285@owczarek.co.uk> <20140408144527.GA5087@netboy> <CALw1_Q1ObS652A+GazmmB4gfkNnnJ4DcsiJ5PoQCd6wxNw-mrA@mail.gmail.com>
Date: Tue, 08 Apr 2014 23:19:02 +0100
Message-ID: <CAHyFsx=nLhD8zn1YdxkEdCOnm=OGTPJf1xD7ER6YhHEpDtJtZQ@mail.gmail.com>
From: Wojciech Owczarek <wojciech@owczarek.co.uk>
To: Kevin Gross <kevin.gross@avanw.com>
Content-Type: multipart/alternative; boundary="001a11337e040cb4ca04f68f5ffc"
Archived-At: http://mailarchive.ietf.org/arch/msg/tictoc/EMTh2UWD7zxEaDVJEtRAD5bHlMc
Cc: "tictoc@ietf.org" <tictoc@ietf.org>
Subject: Re: [TICTOC] Enterprise profile update
X-BeenThere: tictoc@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: Timing over IP Connection and Transfer of Clock BOF <tictoc.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tictoc>, <mailto:tictoc-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/tictoc/>
List-Post: <mailto:tictoc@ietf.org>
List-Help: <mailto:tictoc-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tictoc>, <mailto:tictoc-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 08 Apr 2014 22:19:11 -0000

On 8 April 2014 20:09, Kevin Gross <kevin.gross@avanw.com> wrote:

> The common asymmetric scenario is with PIM-SM. Under PIM-SM, multicasts
> are at least initially routed through a rendezvous point which would likely
> be a detour compared to a unicast path. My understanding is that most
> implementations then go on to build a more reasonable distribution tree for
> multicasts but there is nothing in the PIM RFCs that require or even
> encourage this.
>

There's the RFC and there's best practice. Sane implementations from major
vendors allow you to control the RPT to SPT cutover limits (in bytes,
bandwidth or time), per group or source/group pair, and also on most
enterprise kit (with well developed multicast support), by default, the SPT
cutover happens instantly, as soon as the last hop router receives the
first multicast packet. So if you're the first multicast group joiner, you
probably wait for an announce. Even if the announce is the first packet you
receive, by the time you send your delay request, PIM has already switched
from the shared tree to shortest path tree and multicast is already
following the unicast path.

Again, placing your PIM RPs is a matter of design, a well engineered
network will mitigate the asymmetry on PIM side. If the GM is sufficiently
far away for there to be asymmetries, the traffic is probably going through
your core - and that's where the RP should be. Also you normally do
multiple "instances" of the PIM RP (PIM anycast RPs) in combination with
MSDP so that there is always one close to you.

The case of ECMP remains, but if you want determinism, you design for
determinism.

Where I definitely agree that multicast vs. unicast asymmetry exists, is in
chassis based devices where unicast path in worst case is ASIC-fabric-ASIC,
but multicast is replicated between line cards as a binary tree so latency
can be from unicast latency  to n * unicast latency.

To me, this profile initiative (or standard initiative) is targeted towards
those who specifically need it and a lot of whom already do it.

Regards
Wojciech

-- 
-

Wojciech Owczarek