IMIXS-WORKFLOW Blog

If you migrate form the ix-workflow project to the new Imixs Workflow project it is not possible to redeploy with the new JEE Implementation into an existing EJB / EAR module.

So I recommand to follow these steps:

1.   backup the entities form you existing Imixs Workflow Application with the new (!) org.imixs.workflow.jee.adminclient.web 1.4.3 using the Export Feature.
2. Take care about the new persistence unit used in the persistence.xml
   org.imixs.workflow.jee.jpa
3. Redeploy you new project with the persistence.xml option:
   <property name=”toplink.ddl-generation”
   value=”drop-and-create-tables” />
4. Reimport you backupfile with the new imixs WOrkflow AdminClient Tool

If you have started you workflow project using the build-in Derby Database from Glassfish you maybe can run into a performance problem if you manage a lot and complex of Data. This is because Derby is not so scalable in managing complex SQL statements like other Database Management Systems. In such a case you can easily migrate your workflow system form Derby to MySQL which is much more scalable.

Follow these steps to migrate data in a running system

1. Start your Derby DB and Glassfish Server
2. Make sure that your Workflow System is up and running
3. Switch to the Imixs JEE Admin Client.
4. Connect to the EntityPersistenceManager from your Workflow Instance
5. Start a export using the default EQL Statement. This will export all existing Data of your workflow system into a file
   >SELECT wii FROM Entity wiI
   If the export did not succeed a reason can be the HeapSize of your Glassfish Server instance. You can increase the default HeapSize from 512m to 1024m using the Glassfish Admin Console.
   Go to : Applicationserver->JVM Settings -> JVM Optionsand change -Xmx512m into -Xmx1024m
6. Undeploy your Workflow Instance
7. Shutdown your Derby Database
8. Setup a new MySQL Database and create a new JDBC Connection Pool using the Glassfish Console
9. Now you can switch the JDBC/Ressource of your Workflow System from your Derby Database to your new MySQL Database
10. Deploy your Workflow Instance again – this will generate the necessary Tables in your new MySQL Database
11. Optionall : after deplyoment you can setup Entity Index Fields manually if necessary
12. Import your Data form the Export File you generated in Step 5.

Thats it!

If the import process fails on MySQL check the following:

PACKETTOOBIGEXCEPTION

If you got the Excepiton:

com.mysql.jdbc.PacketTooBigException: Packet for query is too large

you should change the max_allowed_packet setting in your MySQL server for the import process

1.  Open your MySQL console
   >mysql -u root -p
2. Change the max_allowed_packet setting
   >SET GLOBAL max_allowed_packet=3000000000;
3. Restart your Glassfish Server (not the MySQL Database!)
4. Reimport your data
5. Restart you MySQL Server and Glassfish Server to reset the max_allowed_packet settings.
   
   

PERSISTENCEEXCEPTION

If you got the Exeption:

javax.persistence.PersistenceException: Exception [TOPLINK-4002] (Oracle TopLink Essentials – 2.0.1 (Build b09d-fcs (12/06/2007))): oracle.toplink.essentials.exceptions.DatabaseException
Internal Exception: com.mysql.jdbc.MysqlDataTruncation: Data truncation: Data too long for column ‘DATA’ at row 1

MySQL tells you that the BLOB field is to small to store the data. This is because MySQL generates per default a SmallBlob (64K) for your Data Fields in the Entity Table. You can change the Blob type after the deplyoment of your Workflow Instance :

1. Open your MySQL console
   >mysql -u root -p
2. Connect to to your Database
   >USE mydatabase;
3. Change the BLOB Type to LONGBLOB (4GB) or MEDIUMBLOB (16MB)
   >ALTER TABLE Entity CHANGE DATA DATA MEDIUMBLOB NOT NULL;