Mailing-List: contact java-user-help@lucene.apache.org; run by ezmlm
Precedence: bulk
Reply-To: java-user@lucene.apache.org
Received-SPF: pass (nike.apache.org: domain of uwe@thetaphi.de designates
 188.138.97.18 as permitted sender)
From: "Uwe Schindler" <uwe@thetaphi.de>
To: <java-user@lucene.apache.org>
References: <0D638188-E716-4D54-A896-6E3B7511C963@gmail.com>
In-Reply-To: <0D638188-E716-4D54-A896-6E3B7511C963@gmail.com>
Subject: RE: Compound file don't work with lucene 3.3 or 3.4
Date: Sun, 20 Nov 2011 10:44:06 +0100
Message-ID: <00c801cca768$f2438390$d6ca8ab0$@thetaphi.de>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="gb2312"
Content-Transfer-Encoding: quoted-printable
Thread-Index: AQDVyrgVM6o72a9Wn0RyeEsCdFSQAJejOiww
Content-Language: de

Hi,

Lucene only uses CFS for very small segments, that don't exceed a =
specific
percentage of the index size. See the method =
LogMergePolicy.setNoCFSRatio
[http://lucene.apache.org/java/3_4_0/api/all/org/apache/lucene/index/LogM=
erg
ePolicy.html#setNoCFSRatio(double)].

The default is 10%, every segment larger than 0.1 of total index size =
will
not be made compound:
[http://lucene.apache.org/java/3_4_0/api/all/org/apache/lucene/index/LogM=
erg
ePolicy.html#DEFAULT_NO_CFS_RATIO]

In general for merge performance reason you should not change those =
values,
unless you are running out of file descriptors! If you have a static =
index
that never changes (e.g. that will be hosted on a CD-ROM), you can
forceMerge(1) it to one single segment with CFSRatio=3D1.0, in all other =
cases
it's not a good idea, especially if you do realtime indexing; in that =
case
you should use the new TieredMergePolicy not the old LogMergePolicy.

Uwe

-----
Uwe Schindler
H.-H.-Meier-Allee 63, D-28213 Bremen
http://www.thetaphi.de
eMail: uwe@thetaphi.de


> -----Original Message-----
> From: liugang [mailto:liugangc@gmail.com]
> Sent: Sunday, November 20, 2011 5:30 AM
> To: java-user@lucene.apache.org
> Subject: Compound file don't work with lucene 3.3 or 3.4
>=20
> I use lucene 3.4 in my search app.  in default config, after indexing, =
my
index
> dir has several file such as *.tii, *.tis ..., and cfs file doesn't =
exist.
>=20
> Then I use setUseCompoundFile(true) of logByteSizeMergePolicy, and =
result
is
> same.
>=20
> I remember it's good with lucene 2.x.  has anybody same problem?
Appreciate
> for any response.
>=20
> =B7=A2=D7=D4=CE=D2=B5=C4 iPad
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
> For additional commands, e-mail: java-user-help@lucene.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org