Hello everyone, hope you had a great weekend! I am writing this to see if I can get some help. I have found some trouble following an official AWS flink Kinesis Analytics github sample repo about using UDF in a stream. This is the example I am trying to reproduce https://github.com/aws-samples/pyflink-getting-started/tree/main/pyflink-examples/UDF But It's not running. I've tried to see if I have implemented it differently but can't spot the difference. Does someone knows if I am doing something wrong? What I have done so far, is that I've tried doing two scripts: 1.

flink_simple_functional.py

This one does not uses the UDF and just does a very simple transformation like multiplying an integer column by 2 and sinking it to the console. This works fine 2.

flink_simple.py

This other one is a copy of the above mentioned code, but besides doing a multiplication by 2, it also derives another column more by using a UDF. That UDF just converts a string column to lowercase However, the first code works (the one that does not uses the UDF), and the second does not, despite it's a very simple UDF. Can someone spot what might be the mistake? Or if UDF are not supported anymore for AWS Kinesis? I am basically reproducing the AWS repo for flink-kinesis. Here I attach both of my flink files, as well as a third file which is the one I use to populate my Kinesis

input_stream

from which both the flink files fetch as source. Just in case it might help to take a look at how that stream looks like. It only has one shard and I confirm it has data