Logging
Set the transformation logging level ..
Workshop - Logging
Understanding how to effectively use logging is critical for debugging transformations and troubleshooting production issues. Pentaho Data Integration offers multiple logging levels—from "Nothing" to "Row Level"—each providing different amounts of detail about your transformation's execution. Knowing which level to use and how to interpret the output can mean the difference between quickly resolving issues and spending hours hunting for problems.
In this hands-on workshop, you'll build upon your Hello World transformation by exploring PDI's logging capabilities. You'll intentionally introduce a data type mismatch error and then use different logging levels to identify and diagnose the issue. This practical approach mirrors real-world troubleshooting scenarios where understanding log output is essential for maintaining reliable data pipelines.
What You'll Accomplish:
Modify field metadata to create a controlled error scenario
Execute transformations with different logging levels (Basic vs. Row Level)
Interpret the Execution Results and Logging tabs
Identify error messages and stack traces in log output
Understand when to use verbose logging versus minimal logging
Develop troubleshooting skills using PDI's built-in diagnostic tools
By the end of this workshop, you'll understand that while a green checkmark indicates execution completion, it doesn't guarantee success—and you'll know exactly how to use logging to verify your transformations are working correctly. You'll also develop the diagnostic mindset needed to efficiently troubleshoot data integration issues in both development and production environments.
Prerequisites: Completion of the Hello World workshop; Pentaho Data Integration installed and configured
Estimated Time: 5 minutes

Generate Rows
Generate rows outputs a specified number of rows. By default, the rows are empty; however, they can also contain several static fields. This step is used primarily for testing purposes. It may be useful for generating a fixed number of rows, for example, if you require exactly 12 rows for 12 months.
To create an error, change the Type for the field, message, from String to Integer, then view the Execution Results > Logging tab.
Double-click on the Generate Rows Step and change the Type, as illustrated below.

Click OK.
Click the RUN button in the Canvas Toolbar.
Change the Log level from Basic to Row Level.

Click Run. The misconfigured Step is highlighted.

Click on the Log tab in the Execution Results pane.

Its a bit of an art .. but the reason for the error is in there..!

Last updated
Was this helpful?
