Finally figured out the problem.  It turns out that it was a matter of
properly configuring the Geronimo (the application server I use) J2EE
Connector pool settings for the Jackrabbit resource adapter.

The geronimo-ra.xml plan I was using to deploy jackrabbit-jca-1.2.1.rar had
two issues: 1) "thread-caching" was configured for the XA-transactions; and
2) the "single-pool" was being set "match-one".  This caused the same
"connection" to the Jackrabbit repository to be returned.

I am going to close the Jira issue.  For future reference, if you're
deploying Jackrabbit JCA/RAR in Geronimo be sure to configure your Geronimo
deployment plan correctly (refer to:
http://www.chariotsolutions.com/geronimo/geronimo-1.1/connectors.html).

Daniel.

On 2/23/07, Daniel Bloomfield Ramagem < [EMAIL PROTECTED]> wrote:

Done: https://issues.apache.org/jira/browse/JCR-763.

On 2/23/07, Jukka Zitting < [EMAIL PROTECTED]> wrote:
>
> Hi,
>
> On 2/22/07, Daniel Bloomfield Ramagem < [EMAIL PROTECTED]> wrote:
> > Ack, to make matters worse, when I comment out the lines for the
> creation of
> > the second workspace (ws2), the code runs fine but I can't seem to
> login to
> > ws1.  On the file system the "ws1" workspace directory got created
> OK.   But
> > the System.out.println prints "default" for the workspace name)
>
> Sounds alarming. Can you file a bug report for that? Any log files
> with details on what happens within Jackrabbit during the operation
> would be helpful.
>
> BR,
>
> Jukka Zitting
>


Reply via email to