You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@jackrabbit.apache.org by Marcel Reutegger <ma...@gmx.net> on 2009/10/01 11:02:46 UTC

Re: Memory issues with jackrabbit/lucene

Hi,

On Fri, Sep 25, 2009 at 14:24, Muguet Bradbury <M....@fortent.com> wrote:
> Is there a way to add content to a jackrabbit repo without making it
> searchable, thus avoiding the TermBuffer objects being created?

you can define indexing rules for certain node types and let
jackrabbit only index certain properties.

see: http://wiki.apache.org/jackrabbit/IndexingConfiguration

regards
 marcel

RE: Memory issues with jackrabbit/lucene

Posted by Muguet Bradbury <M....@fortent.com>.
I wanted to let everyone know how we solved this memory issue with our
jackrabbit repository.

As was stated earlier in the thread, previously we stored all of our
large documents as custom properties on the jackrabbit node.  We did not
use jcr:mimeType or jcr:data.  The indexing of these string properties
were filling up the memory, and the CPU was thrashing attempting to keep
up with indexing on these documents.

The change we made was to store the documents in the jcr:data property
of a leaf node, with a jcr:mimeType of application/octet-stream.  This
turned off the indexing and our application is much happier now.

Thanks for all your help.

-----Original Message-----
From: Alexander Klimetschek [mailto:aklimets@day.com] 
Sent: Thursday, October 01, 2009 8:48 AM
To: users@jackrabbit.apache.org
Subject: Re: Memory issues with jackrabbit/lucene

Yes, wiki.apache.org is down. http://monitoring.apache.org/status/
gives red for eos.apache.org. So we all have to wait a moment until
it's up again.

Regards,
Alex

-- 
Alexander Klimetschek
alexander.klimetschek@day.com

______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
______________________________________________________________________

Re: Memory issues with jackrabbit/lucene

Posted by Alexander Klimetschek <ak...@day.com>.
Yes, wiki.apache.org is down. http://monitoring.apache.org/status/
gives red for eos.apache.org. So we all have to wait a moment until
it's up again.

Regards,
Alex

-- 
Alexander Klimetschek
alexander.klimetschek@day.com

RE: Memory issues with jackrabbit/lucene

Posted by Muguet Bradbury <M....@fortent.com>.
Marcel,

I'm very interested in this solution, but the link you gave does not
work.

-----Original Message-----
From: mreutegg@day.com [mailto:mreutegg@day.com] On Behalf Of Marcel
Reutegger
Sent: Thursday, October 01, 2009 5:03 AM
To: users@jackrabbit.apache.org
Subject: Re: Memory issues with jackrabbit/lucene

Hi,

On Fri, Sep 25, 2009 at 14:24, Muguet Bradbury <M....@fortent.com>
wrote:
> Is there a way to add content to a jackrabbit repo without making it
> searchable, thus avoiding the TermBuffer objects being created?

you can define indexing rules for certain node types and let
jackrabbit only index certain properties.

see: http://wiki.apache.org/jackrabbit/IndexingConfiguration

regards
 marcel

______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
______________________________________________________________________