How to use notifications and logs to spot and correct errors in a pipeline which
execution has failed.
Before you begin
You have previously created a complete pipeline.
You have executed your pipeline and the execution failed.
Procedure
Open the pipeline that failed.
Here, this pipeline contains a database table about
customers, a Python 3
processor that concatenate names, converts currency
and processes dates, and a HDFS topic to store the
processed data.
To display the error message that informs you about the execution status, click
the icon on the top
toolbar.
From that window, you can expand the error
message by clicking on it.
To get more detailed and categorized information about your
error, either click the View Logs
link from the Notification window, or go to the
Metrics tab of the
Pipeline Details panel and
click the View Logs button.
By default, all logs are displayed.
To sort the logs and only read the error messages, select the
Error check box.
You can see that the error is linked to an
incorrect operator used in the Python
3 processor on line 7.
Now that you have spotted the error, select the
Python 3 processor to edit
its configuration.
On line 7 here, the year 2017 is surrounded by
unnecessary operators that cause the pipeline to
fail. Edit the line as
follows:output['number_year_registrated']
= 2017 - int(year) and save your
changes.
Click the run icon on the top toolbar to execute your pipeline
again.
Results
The pipeline is fixed and is executed with success. If you missed the notification
informing about the execution success, click the icon on the top toolbar to display it.
Did this page help you?
If you find any issues with this page or its content – a typo, a missing step, or a technical error – please let us know!