Mailing-List: contact users-help@jackrabbit.apache.org; run by ezmlm
Precedence: bulk
Reply-To: users@jackrabbit.apache.org
Received-SPF: pass (asf.osuosl.org: domain of marcel.reutegger@gmx.net
 designates 213.165.64.20 as permitted sender)
Message-ID: <44E57300.7030408@gmx.net>
Date: Fri, 18 Aug 2006 09:57:52 +0200
From: Marcel Reutegger <marcel.reutegger@gmx.net>
User-Agent: Thunderbird 1.5 (Windows/20051201)
MIME-Version: 1.0
To: users@jackrabbit.apache.org
Subject: Re: Optimizing the index
References: <5860763.post@talk.nabble.com>
In-Reply-To: <5860763.post@talk.nabble.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

The index in Jackrabbit will optimize itself based on configuration 
properties. There is currently no method you can call to optimize the 
index manually. If you think this is a needful enhancement can you 
please create a jira issue: http://issues.apache.org/jira/browse/JCR

Per default index folders with 100 nodes are created initially. When 
there are 10 index folders they are merged and optimized into a single 
index folder with approximately 1000 (=10*100) nodes. Similarly when 
there are 10 index folders with each 1000 nodes those are merged and 
optimized into a single one. The resulting index folder will then have 
approximately 10000 nodes. The maximum number of nodes merged and 
optimized into a single index is controlled by the 'maxMergeDocs' 
parameter. The default value is 100'000. For your estimated number of 
documents this is too low. You should increase this value to at least 
1'000'000.

For a quick intro to the inner workings of the query engine see:
http://jackrabbit.apache.org/doc/arch/operate/query.html

See also the 'SearchIndex' section in:
http://svn.apache.org/repos/asf/jackrabbit/trunk/jackrabbit/src/main/config/repository.xml

regards
  marcel

sowmi wrote:
> How do I optimize the index that I am creating? Are there hooks in JackRabbit
> to let me trigger an optimization? Right now, I have an index of around 120
> MB size with 400k documents in it. My ${repository}/workspaces/default/index
> directory has around 19 folders. Can I optimize this somehow, as my total
> document size when I am done will be around 10 million. Please advise.
> 
> sowmi