You config looks good. I think it would be great to update the docs. Can
you edit this yourself in the pax-jdbc wiki?
If you do not have wiki access you can ask for it on the ops4j mailing list.
The error you get: Auto-commit can not be set while enrolled in a
transaction" might be an issue in dbcp2. I thought it would be solved
already.
You can try to report it at openjpa and dbcp2. If you do not use the
newest pax-jdbc version then you could also try that.
Unfortunately I was not able to find the issue at pax-jdbc or dbcp2.
Christian
On 11.07.2016 10:00, Bengt Rodehav wrote:
Thanks Christian,
It seemed the reason why the pax-jdbc datasource wasn't picked up by
JPA Blueprint was that I had given it the wrong name...
I had forgotten the "jdbc/" prefix that I use in my persistence.xml.
It is now picked up as it should.
There is, however, an error in the documentation in the link you gave
me. It says:
---
Method Arguments
The argument passed to |createDataSource(),
createConnectionPoolDataSource(), createXADataSource()| supports the
following properties:
* |DataSourceFactory.JDBC_DATABASE_NAME| (mandatory)
* |DataSourceFactory.JDBC_USER|
* |DataSourceFactory.JDBC_PASSWORD|
_An |SQLException| is thrown if any other properties are set_ or if a
mandatory property is missing.
--
But I need to pass properties directly to the jdbc driver. In my case
"createDatabase=create" so that the database is created if it doesn't
already exists. I therefore set this property in my
org.ops4j.datasource configuration as follows:
osgi.jdbc.driver.name <http://osgi.jdbc.driver.name>=derby-pool-xa
databaseName=historyDB
dataSourceName=jdbc/filetransferhistory
createDatabase=create
The "createDatabase" property is passed to the driver so that the
database is created. So, it is OK to pass other properties as well -
they get passed along.
However, I do still have a problem getting the Derby database to be
created. When trying to create a new database I get:
Caused by: java.sql.SQLException: Auto-commit can not be set while
enrolled in a transaction
at
org.apache.commons.dbcp2.managed.ManagedConnection.setAutoCommit(ManagedConnection.java:223)
at
org.apache.openjpa.lib.jdbc.DelegatingConnection.setAutoCommit(DelegatingConnection.java:167)
at
org.apache.openjpa.lib.jdbc.DelegatingConnection.setAutoCommit(DelegatingConnection.java:167)
at
org.apache.openjpa.lib.jdbc.ConfiguringConnectionDecorator$ConfiguringConnection.setAutoCommit(ConfiguringConnectionDecorator.java:116)
at
org.apache.openjpa.jdbc.schema.SchemaTool.executeSQL(SchemaTool.java:1297)
at
org.apache.openjpa.jdbc.schema.SchemaTool.createTable(SchemaTool.java:1017)
at
org.apache.openjpa.jdbc.schema.SchemaTool.buildSchema(SchemaTool.java:573)
at
org.apache.openjpa.jdbc.schema.SchemaTool.add(SchemaTool.java:481)
at
org.apache.openjpa.jdbc.schema.SchemaTool.add(SchemaTool.java:368)
at
org.apache.openjpa.jdbc.schema.SchemaTool.run(SchemaTool.java:343)
at
org.apache.openjpa.jdbc.meta.MappingTool.record(MappingTool.java:507)
... 46 more
It works if I instead use "derby" or "derby-pool" DataSourceFactory. I
tried to provide a <non-jta-datasource> for this purpose but it is
never being used. I thought this was the reason why a
<non-jta-datasource> should be provided. Not sure whether this is a
problem with JPA blueprint or if its a problem in OpenJPA. Maybe I
should ask on the OpenJPA mailing list?
/Bengt
--
Christian Schneider
http://www.liquid-reality.de
Open Source Architect
http://www.talend.com
