This post continues on my series of best practices around the installation and configuration of Pentaho's Data Integration (PDI) tool suit and addresses JNDI configuration of database connections.
JNDIs' contribution to more robust data projects stems from the fact that offers a central place to configure all your database connections. This is far more superior than using JDBC connections defined in every one of your transformation files within your project. It serves to minimise the impact any database environmental or account changes. The PDI has JNDI support through the simple-jndi library.
The original post that triggered me to write this post written by the kettle-pentaho blogger site. If you have a Java software engineering background, then most of this will be very familiar to you. The audience I intend to help with these posts however, are more likely not to have that experience. So I intend to flesh out the description to make it easier for folks who are neither software nor data engineers.
I will cover using the PDI JNDI in three parts:
- The first part covers configuring the PDI JNDI properties
- The second part describes how to configure a JNDI connection within Spoon and finally
- The third part address how to troubleshoot a PDI JNDI configuration that is not working
Note that this post has assumed you've already downloaded the database vendor's JDBC jar file and copied it to the
pdi-install-directory/lib directory. Of course, it also assumes you have the database connection details to which you wish to connect (including the server's IP or DNS, the database port, database name and the account details).
I hope some readers will get some benefit from the information provided. Please comment on my discourse site if the material benefited you.