Re: [hybi] Last Call: <draft-ietf-hybi-thewebsocketprotocol-10.txt>

Martin Rex <> Wed, 27 July 2011 23:51 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id 044DA21F8A7D; Wed, 27 Jul 2011 16:51:18 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -9.902
X-Spam-Status: No, score=-9.902 tagged_above=-999 required=5 tests=[AWL=0.347, BAYES_00=-2.599, HELO_EQ_DE=0.35, RCVD_IN_DNSWL_HI=-8]
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id hwEL2AaTBIoi; Wed, 27 Jul 2011 16:51:17 -0700 (PDT)
Received: from ( []) by (Postfix) with ESMTP id 348DD21F8A64; Wed, 27 Jul 2011 16:51:17 -0700 (PDT)
Received: from by (26) with ESMTP id p6RNoeDx027467 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK); Thu, 28 Jul 2011 01:50:40 +0200 (MEST)
From: Martin Rex <>
Message-Id: <>
To: (Mark Andrews)
Date: Thu, 28 Jul 2011 01:50:39 +0200 (MEST)
In-Reply-To: <> from "Mark Andrews" at Jul 28, 11 09:09:38 am
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
X-SAP: out
X-Mailman-Approved-At: Wed, 27 Jul 2011 20:40:52 -0700
Subject: Re: [hybi] Last Call: <draft-ietf-hybi-thewebsocketprotocol-10.txt>
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: Server-Initiated HTTP <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Wed, 27 Jul 2011 23:51:18 -0000

Mark Andrews wrote:
> Dave Cridland writes:
> > 
> > Happy eyeballs - try everything as soon as you can, in parallel. Drop  
> > everything else when one does.
> More correctly it is try the first address and if that doesn't
> connect in a short period (150...250ms) start a second connection
> to the next address while continuing with the first.  If you have
> more that 2 address you do something similar for the next one (I
> use 1/2 the original timeout, but that is a implementation detail).
> You continue to use the address that works for that session.  You
> drop any other connections to other addresses that complete.

Happy eyeballs means that a clients reaction to congestion is
to perform an DoS attack, flood the network with additional
connection requests and hammer the server with many additional
half-open connections that will never actually get used.

While this might currently "improve" the end user experience
of clients, it simultaneously adds a deterrant to server operators
to announce IPv6 addresses (even multiple IP addresses -- they're
better of with IPv4 NAT if they have multiple servers at a single