Mailing-List: contact user-help@couchdb.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@couchdb.apache.org
Received-SPF: pass (athena.apache.org: domain of adam.kocoloski@gmail.com
 designates 209.85.210.52 as permitted sender)
Content-Type: text/plain; charset=us-ascii
Mime-Version: 1.0 (Apple Message framework v1081)
Subject: Re: Questions about database file sizes.
From: Adam Kocoloski <kocolosk@apache.org>
In-Reply-To: <AANLkTimBTH4W58aU7t+L+XdMuHfMHGnvFtHbk3JJjt91@mail.gmail.com>
Date: Mon, 8 Nov 2010 09:03:09 -0500
Content-Transfer-Encoding: quoted-printable
Message-Id: <037AF5C1-3DF5-4F0A-A4A5-966882FC2C7A@apache.org>
References: <AANLkTinCyXhW3i0A13G+rKO7_J6wNBBp+YbVuaL5ro+e@mail.gmail.com>
 <AANLkTinrfVOQByXuD9Vi=TirWZW2kMLTiKuB5MH0YawQ@mail.gmail.com>
 <AANLkTimPypVR4qQGANcegY7Xg0sO8s5_u3gn8579iyP9@mail.gmail.com>
 <AANLkTimBTH4W58aU7t+L+XdMuHfMHGnvFtHbk3JJjt91@mail.gmail.com>
To: user@couchdb.apache.org

Replication won't copy the _local docs between the nodes.  You might =
have some of those on only one node, e.g. from a replication that only =
involved one of the two databases.  Regards,

Adam

On Nov 7, 2010, at 3:31 PM, Sivan Greenberg wrote:

> So far, I get the differences even after compacting twice and three =
times.
>=20
> On Sun, Nov 7, 2010 at 10:21 PM, Paul Davis =
<paul.joseph.davis@gmail.com> wrote:
>> On Sun, Nov 7, 2010 at 12:02 PM, Sivan Greenberg <sivang@gmail.com> =
wrote:
>>> ---------- Forwarded message ----------
>>> From: Sivan Greenberg <sivan@omniqueue.com>
>>> Date: Sun, Nov 7, 2010 at 6:58 PM
>>> Subject: Questions about database file sizes.
>>> To: users@couchdb.apache.org
>>>=20
>>>=20
>>> Hi List,
>>>=20
>>>  First I'd like to say that my longly developed couchdb based system
>>> seems to be performing well and serving its purpose.
>>>=20
>>>  However, there's always a small difference between the database =
files
>>> on the 2 couchdb server nodes we are using, after identical =
compaction
>>> steps on both db's, the difference remains around 4k, even though =
the
>>> number of documents is the same.
>>>=20
>>>  This is CouchDB 0.11.0 , 2 nodes, each one does a pull replication
>>> from its peer in continue replication over HTTP.
>>>=20
>>>  Is this normal? Can we sleep at night without fearing the delta =
would
>>> at some edge case occasion increase to the gigs ? Is it unrealistic =
to
>>> expect the two files to be of the same size given same compaction
>>> steps were taken and they both replicate pull from each other?
>>>=20
>>> Many thanks,
>>> (CouchDB Rocks!)
>>>=20
>>> -Sivan
>>>=20
>>=20
>> You might try compacting twice. If you're not seeing the difference
>> change after continuous writes its probably not a big deal.
>>=20
>> Paul
>>=20