Received-SPF: pass (sog-mx-3.v43.ch3.sourceforge.com: domain of gmail.com
	designates 209.85.220.175 as permitted sender)
	client-ip=209.85.220.175; envelope-from=mh.in.england@gmail.com;
	helo=mail-vx0-f175.google.com; 
MIME-Version: 1.0
Sender: mh.in.england@gmail.com
In-Reply-To: <4E662B79.5090303@gmail.com>
References: <4E65CEE6.7030002@gmail.com> <4E65DA06.9060403@gmail.com>
	<CALxbBHUajARXc1oA-NjD+U8hW5uSqF=u4ZHHBfcmT_O8GjpNiA@mail.gmail.com>
	<CANEZrP0VXDUs_mAKCVKD1Q0ijyb989oADrCN1zTZ1nnN_JQ=cQ@mail.gmail.com>
	<4E661FAE.9020008@gmail.com>
	<CANEZrP3=UPYkBQo6b421xaMGyP4BsGiw8DBuM8pT2ow1Vom9JQ@mail.gmail.com>
	<4E662B79.5090303@gmail.com>
Date: Tue, 6 Sep 2011 16:52:20 +0200
Message-ID: <CANEZrP2Wh82sqGjZDn_M=UPufBCU4fP9zEXV_K8JpgVF8O1FCw@mail.gmail.com>
From: Mike Hearn <mike@plan99.net>
To: shadders.del@gmail.com
Content-Type: multipart/alternative; boundary=bcaec51d2780872fd304ac46f916
Cc: bitcoin-development@lists.sourceforge.net
Subject: Re: [Bitcoin-development] Building a node crawler to map network
Precedence: list

--bcaec51d2780872fd304ac46f916
Content-Type: text/plain; charset=UTF-8

On Tue, Sep 6, 2011 at 4:17 PM, Steve <shadders.del@gmail.com> wrote:

> **
> I'm not really understanding the use case though.  I believe most
> bitcoind's have a default max connections of 8.  Is the goal to increase
> this without fundamentally altering the bitcoind concurrency model?
>

bitcoind already uses asynchronous IO. That's not the problem.

The issue came up in a conversation about scalability. If Bitcoins
popularity continues to grow, users are very likely to migrate away from
running full verifying nodes to lightweight clients, either a different mode
of the Satoshi client or different implementations like the Android Wallet
or MultiBit.

Lightweight clients cannot verify thus should not relay. And they'll be run
by users who just want to send/receive coins from time to time, so don't
leave the programs running 24/7. The result could be running out of sockets
(like we have had problems with recently). It's especially true because
lightweight clients cannot check transactions for themselves. If they want
to show transactions appearing immediately (and they do), they have to use
"heard from lots of nodes" as a proxy for validity. So lightweight clients
are likely to be socket intensive.

We could solve this by just hoping that lots of people run full nodes. The
problem is that a full node is quite an intensive thing already, it uses
lots of CPU and disk seeks, and will just get more expensive in future. And
as transaction traffic increases, that leaves less CPU time available to
service thousands of connected clients. The ROI of bringing up a new node
decreases at the same time as the userbase increases.

One traditional approach to solving this is frontend proxies. Jabber.com/org
used this technique many years ago, and Google has also used it to scale up the
lockservice<http://static.googleusercontent.com/external_content/untrusted_dlcp/labs.google.com/en/us/papers/chubby-osdi06.pdf>
(see
section 3.1). It's effective because often maintaining connections to
thousands of clients doesn't involve much brainwork, just shifting bytes
around. This is especially true of Bitcoin. So if somebody is running a full
node already they could increase their client capacity by just bringing up a
frontend proxy and having it handle things like outbound tx
broadcasts/deduping inbound broadcasts, connection setup, relaying recently
found blocks etc. A well written proxy could probably support tens of
thousands of simultaneous clients which frees up the bitcoinds time for
verification and wallet manipulation.

--bcaec51d2780872fd304ac46f916
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div class=3D"gmail_quote">On Tue, Sep 6, 2011 at 4:17 PM, Steve <span dir=
=3D"ltr">&lt;<a href=3D"mailto:shadders.del@gmail.com">shadders.del@gmail.c=
om</a>&gt;</span> wrote:<br><blockquote class=3D"gmail_quote" style=3D"marg=
in:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<u></u>

 =20
   =20
 =20
  <div bgcolor=3D"#ffffff" text=3D"#000000">I&#39;m not really understandin=
g the use case though.=C2=A0 I believe most
    bitcoind&#39;s have a default max connections of 8.=C2=A0 Is the goal t=
o
    increase this without fundamentally altering the bitcoind
    concurrency model?</div></blockquote><div><br></div><div>bitcoind alrea=
dy uses asynchronous IO. That&#39;s not the problem.</div><div><br></div><d=
iv>The issue came up in a conversation about scalability. If Bitcoins popul=
arity continues to grow, users are very likely to migrate away from running=
 full verifying nodes to lightweight clients, either a different mode of th=
e Satoshi client or different implementations like the Android Wallet or Mu=
ltiBit.</div>
<div><br></div><div>Lightweight clients cannot verify thus should not relay=
. And they&#39;ll be run by users who just want to send/receive coins from =
time to time, so don&#39;t leave the programs running 24/7. The result coul=
d be running out of sockets (like we have had problems with recently). It&#=
39;s especially true because lightweight clients cannot check transactions =
for themselves. If they want to show transactions appearing immediately (an=
d they do), they have to use &quot;heard from lots of nodes&quot; as a prox=
y for validity. So lightweight clients are likely to be socket intensive.</=
div>
<div><br></div><div>We could solve this by just hoping that lots of people =
run full nodes. The problem is that a full node is quite an intensive thing=
 already, it uses lots of CPU and disk seeks, and will just get more expens=
ive in future. And as transaction traffic increases, that leaves less CPU t=
ime available to service thousands of connected clients. The ROI of bringin=
g up a new node decreases at the same time as the userbase increases.</div>
<div><br></div><div>One traditional approach to solving this is frontend pr=
oxies. Jabber.com/org used this technique many years ago, and Google has al=
so used it to scale up <a href=3D"http://static.googleusercontent.com/exter=
nal_content/untrusted_dlcp/labs.google.com/en/us/papers/chubby-osdi06.pdf">=
the lockservice</a>=C2=A0(see section 3.1). It&#39;s effective because ofte=
n maintaining connections to thousands of clients doesn&#39;t involve much =
brainwork, just shifting bytes around. This is especially true of Bitcoin. =
So if somebody is running a full node already they could increase their cli=
ent capacity by just bringing up a frontend proxy and having it handle thin=
gs like outbound tx broadcasts/deduping inbound broadcasts, connection setu=
p, relaying recently found blocks etc. A well written proxy could probably =
support tens of thousands of simultaneous clients which frees up the bitcoi=
nds time for verification and wallet manipulation.</div>
</div>

--bcaec51d2780872fd304ac46f916--