You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@jackrabbit.apache.org by Pedro Alondra <pe...@yahoo.es> on 2009/05/19 12:25:04 UTC

multiple clients vs jackrabbit repository

Hi All,

I have a problem with jackrabbit repository being accessed from multiple clients.
I would like to initialize the repository using command line tool using
repository.xml configured to use database via JDBC as the persistence storage.
Then, I would like to access this repository from within web application
using JNDI-based repository.xml (JNDI pointing at the same database).

It seems there is a problem seeing all the workspaces created by command line 
tool from within the web application (and vice versa).

Do you have any suggestions how do I make it?

Regards,
Pedro



      

Re: multiple clients vs jackrabbit repository

Posted by Alexander Klimetschek <ak...@day.com>.
On Wed, May 20, 2009 at 4:34 AM, Pedro Alondra <pe...@yahoo.es> wrote:
> In command line tool (that uses JDBC-based repository.xml) I can see only the workspaces created by itself,
> I cannot see any workspaces that were created by webapp client (that uses JNDI-based repository.xml).
>
> It seems that workspaces information is kept in repository directory, and if both clients
> don't share one directory, then they cannot see eachothers workspaces. If I define
> one shared repository directory, then workspace.xml is going to contain JNDI/JDBC configuration,
> that becomes irrelevant for one of the actors (JNDI vs JDBC).

This looks like you have actually to repositories, ie. two different
repository home directories.

The list of workspaces is defined by the list of subdirectories of
<repository-home>/workspaces (or whatever is set as "rootPath"
attribute on the <Workspaces /> element in repository.xml). Since each
workspace can use a completely different persistence manager, eg. you
could have one using some file-system based storage and another one
using a database, this information is stored in the filesystem and not
the database you happen to configure for the storage of one or more
workspaces.

Regards,
Alex

-- 
Alexander Klimetschek
alexander.klimetschek@day.com

Re: multiple clients vs jackrabbit repository

Posted by Pedro Alondra <pe...@yahoo.es>.


Alexander Klimetschek wrote:

> On Tue, May 19, 2009 at 5:25 AM, Pedro Alondra <pe...@yahoo.es> wrote:
>> It seems there is a problem seeing all the workspaces created by command line
>> tool from within the web application (and vice versa).
>
> You can normally only see (= list) all workspaces
> (session.getWorkspace().getAccessibleWorkspaceNames()) if you are
> logged in as an "admin" user.

In command line tool (that uses JDBC-based repository.xml) I can see only the workspaces created by itself,  
I cannot see any workspaces that were created by webapp client (that uses JNDI-based repository.xml).

It seems that workspaces information is kept in repository directory, and if both clients
don't share one directory, then they cannot see eachothers workspaces. If I define
one shared repository directory, then workspace.xml is going to contain JNDI/JDBC configuration,
that becomes irrelevant for one of the actors (JNDI vs JDBC).

So, the question is, how is it possible to have two clients using different ways of accessing
the same database storage and that they'll be able to see eachother workspaces?

Regards,
Pedro



      

Re: multiple clients vs jackrabbit repository

Posted by Alexander Klimetschek <ak...@day.com>.
On Tue, May 19, 2009 at 5:25 AM, Pedro Alondra <pe...@yahoo.es> wrote:
> It seems there is a problem seeing all the workspaces created by command line
> tool from within the web application (and vice versa).

You can normally only see (= list) all workspaces
(session.getWorkspace().getAccessibleWorkspaceNames()) if you are
logged in as an "admin" user.

Regards,
Alex

--
Alexander Klimetschek
alexander.klimetschek@day.com