Review the rules and guidelines for configuring a mapping in advanced mode.
The Source transformation returns a JSON parsing error, even though my JSON file is good. What should I do?
When you select a JSON file in the source object, you receive the following error:
error [The file or partition directory [<path>/<JSON file>] is not valid. The parser encountere the following error while parsing the content: [Unexpected character 1 at 1]. Select a valid [Json] file or partition director.]
The error occurs when there is an array of primitives without a field name in the JSON file. For example, the following array causes a parsing error: [10,20,30,40,50,0]
The Source transformation parses the following array successfully: {"numbers":[10,20,30,40,50,0]}
As a workaround, wrap the top-level array in a struct element before reading the file in the Source transformation.
I received the "version 'GLIBC_<version>' not found" error when running a mapping in advanced mode
This error occurs when you install Python on a RHEL 9.x-compatible Secure Agent machine and then use the Python installation in a mapping in advanced mode.
The mapping in advanced mode fails because the custom Python installation requires a higher version of GLIBC than what's in the Spark Docker image.
To resolve the error, locate a Secure Agent machine still running RHEL 8.x and install Python on that machine. Then, copy the installation folder to the RHEL 9.x machine.
Note: If you're migrating an existing agent node from a RHEL 7.x or 8.x compatible machine to a RHEL 9.x compatible machine, copy the Python installation folder from the older agent node to the new agent node.