jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Alexander Klimetschek" <aklim...@day.com>
Subject Re: Filesystem, datastore and mysql in JackRabbit
Date Fri, 10 Oct 2008 11:18:02 GMT
Basically all data (JCR nodes and properties) is stored in the
PersistenceManager, eg. the mysql db in your case. The FileSystem is
only used if the persistence manager uses it, which is not the case
for the (recommended) bundle db persistence managers. The Lucene
search index always uses the filesystem
(workspaces/WORKSPACE_NAME/index). The datastore is used for large
binary properties (eg. files stored as nt:file in the repo), which are
faster when stored separately from smaller properties and on the
filesystem than in a database (and it avoids duplicates, see
http://wiki.apache.org/jackrabbit/DataStore).

Regards,
Alex

On Fri, Oct 10, 2008 at 12:44 PM, RafalJanik <rajn@softwaremind.pl> wrote:
> Hi, I'm trying to understand how JacRabbit works...
> In my repository.xml I have <datastore> <persistanceManager> with mysql.
> In datastore there is sended date which size is lower then 100.
> I know that in MySQL db JackRabbit creates four tables for workspace, but
> when add nodes with the same properties only the workspace in filesystem is
> changing and tables in mysqldb. My question is what exactly is putting into
> db and what into filesystem after adding a new node? Is workspace filesystem
> only working copy of part of data stored in db? How works connections
> between datastore and db?
>
> Cheers,
>
> Rafael
>
>
>



-- 
Alexander Klimetschek
alexander.klimetschek@day.com

Mime
View raw message