Received-SPF: pass (sog-mx-2.v43.ch3.sourceforge.com: domain of gmail.com
	designates 209.85.214.43 as permitted sender)
	client-ip=209.85.214.43; envelope-from=mh.in.england@gmail.com;
	helo=mail-bk0-f43.google.com; 
MIME-Version: 1.0
Sender: mh.in.england@gmail.com
In-Reply-To: <CAC1+kJO8KX6uc7jEz9Otv+BbrJQUk6RvbiyWnP0m=Voh8zdcWQ@mail.gmail.com>
References: <CANEZrP1ksdGfB-282giysF-2Lo==NiD-nBdon5FJ9wLBLmX=TA@mail.gmail.com>
	<CAC1+kJO8KX6uc7jEz9Otv+BbrJQUk6RvbiyWnP0m=Voh8zdcWQ@mail.gmail.com>
Date: Tue, 17 Sep 2013 13:45:18 +0200
Message-ID: <CANEZrP2ZRHfoGugVZ2X9iE8vzdZydJ3JTS=roTv8G=LnCgm7Eg@mail.gmail.com>
From: Mike Hearn <mike@plan99.net>
To: =?UTF-8?B?Sm9yZ2UgVGltw7Nu?= <jtimon@monetize.io>
Content-Type: multipart/alternative; boundary=485b3970d514e6ac1404e692daa2
Cc: Bitcoin Dev <bitcoin-development@lists.sourceforge.net>
Subject: Re: [Bitcoin-development] Faster databases than LevelDB
Precedence: list

--485b3970d514e6ac1404e692daa2
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Nobody has written code to use a better format, migrate old wallets, etc.


On Tue, Sep 17, 2013 at 1:41 PM, Jorge Tim=C3=B3n <jtimon@monetize.io> wrot=
e:

> Only slightly related to this...
> What's the reason why BerkleyDB is maintained for the wallet?
> I think it would be a good thing to get rid of the libdb4.8++-dev
> dependency that makes bitcoind harder to compile on debian and ubuntu.
> Unless, of course, there's a reason I am missing...
>
>
> On 9/17/13, Mike Hearn <mike@plan99.net> wrote:
> > LevelDB is fast - very fast if you give it enough CPU time and disk
> seeks.
> > But it's not the last word in performance.
> >
> > HyperLevelDB is a forked LevelDB with some changes, mostly, finer grain=
ed
> > locking and changes to how compaction works:
> >
> > http://hyperdex.org/performance/leveldb/
> >
> > However, it comes with a caveat - one of the changes they made is to ta=
ke
> > away write throttling if compaction falls behind, the app itself is
> > expected to do that.
> >
> > Sophia is a competitor to LevelDB. The website claims that in benchmark=
s
> it
> > completely smokes LevelDB. I have not explored how it does this or trie=
d
> to
> > replicate their benchmarks myself:
> >
> > http://sphia.org/index.html
> > http://sphia.org/benchmarks.html
> >
> > It's written in C and BSD licensed.
> >
> > As an example of the kind of speedup they claim to be capable of, they
> say
> > LevelDB could do 167,476 random reads per second on their SSD based
> > machine. Sophia could do 438,084 reads/sec. Random reads are of course
> the
> > most interesting for us because that's what UTXO lookups involve.
> >
> > They also compare against HyperLevelDB, where the differences are much
> less
> > pronounced and actually HyperLevelDB appears to be able to do random
> writes
> > faster than Sophia.
> >
>
>
> --
> Jorge Tim=C3=B3n
>
> http://freico.in/
>

--485b3970d514e6ac1404e692daa2
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Nobody has written code to use a better format, migrate ol=
d wallets, etc.</div><div class=3D"gmail_extra"><br><br><div class=3D"gmail=
_quote">On Tue, Sep 17, 2013 at 1:41 PM, Jorge Tim=C3=B3n <span dir=3D"ltr"=
>&lt;<a href=3D"mailto:jtimon@monetize.io" target=3D"_blank">jtimon@monetiz=
e.io</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">Only slightly related to this...<br>
What&#39;s the reason why BerkleyDB is maintained for the wallet?<br>
I think it would be a good thing to get rid of the libdb4.8++-dev<br>
dependency that makes bitcoind harder to compile on debian and ubuntu.<br>
Unless, of course, there&#39;s a reason I am missing...<br>
<div class=3D"HOEnZb"><div class=3D"h5"><br>
<br>
On 9/17/13, Mike Hearn &lt;<a href=3D"mailto:mike@plan99.net">mike@plan99.n=
et</a>&gt; wrote:<br>
&gt; LevelDB is fast - very fast if you give it enough CPU time and disk se=
eks.<br>
&gt; But it&#39;s not the last word in performance.<br>
&gt;<br>
&gt; HyperLevelDB is a forked LevelDB with some changes, mostly, finer grai=
ned<br>
&gt; locking and changes to how compaction works:<br>
&gt;<br>
&gt; <a href=3D"http://hyperdex.org/performance/leveldb/" target=3D"_blank"=
>http://hyperdex.org/performance/leveldb/</a><br>
&gt;<br>
&gt; However, it comes with a caveat - one of the changes they made is to t=
ake<br>
&gt; away write throttling if compaction falls behind, the app itself is<br=
>
&gt; expected to do that.<br>
&gt;<br>
&gt; Sophia is a competitor to LevelDB. The website claims that in benchmar=
ks it<br>
&gt; completely smokes LevelDB. I have not explored how it does this or tri=
ed to<br>
&gt; replicate their benchmarks myself:<br>
&gt;<br>
&gt; <a href=3D"http://sphia.org/index.html" target=3D"_blank">http://sphia=
.org/index.html</a><br>
&gt; <a href=3D"http://sphia.org/benchmarks.html" target=3D"_blank">http://=
sphia.org/benchmarks.html</a><br>
&gt;<br>
&gt; It&#39;s written in C and BSD licensed.<br>
&gt;<br>
&gt; As an example of the kind of speedup they claim to be capable of, they=
 say<br>
&gt; LevelDB could do 167,476 random reads per second on their SSD based<br=
>
&gt; machine. Sophia could do 438,084 reads/sec. Random reads are of course=
 the<br>
&gt; most interesting for us because that&#39;s what UTXO lookups involve.<=
br>
&gt;<br>
&gt; They also compare against HyperLevelDB, where the differences are much=
 less<br>
&gt; pronounced and actually HyperLevelDB appears to be able to do random w=
rites<br>
&gt; faster than Sophia.<br>
&gt;<br>
<br>
<br>
</div></div><span class=3D"HOEnZb"><font color=3D"#888888">--<br>
Jorge Tim=C3=B3n<br>
<br>
<a href=3D"http://freico.in/" target=3D"_blank">http://freico.in/</a><br>
</font></span></blockquote></div><br></div>

--485b3970d514e6ac1404e692daa2--