Execution never ends with a bad Redshift connection

[rafaelleinio]

I experienced this issue testing aws-data-wrangler in the "Spark to Redshift" use case.

I created a Redshift connection with the method Redshift.generate_connection and when I tried to load the dataframe to Redshift with the method session.spark.to_redshift the executions never finished so I had to manually cancel the execution. The temporary files on S3 were created, but nothing on Redshift. Later I realize that I was passing the port as string to the connection but I think that the method expects an int argument, when I changed this the load worked just fine!

My guess is that with a wrong connection setup wrangler can't reach Redshift but there is no timeout to stop the execution, so the code gets stuck.

[igorborgest]

Thanks @rafaelleinio!

I solved the port typing issue, and also created a more generic connection validation mechanism.

Make sense?