CSV - convert to file Node, Bug: 1.45.0

I have a workflow, where I do some ETL.

Load from Postgres, filter, add/transform some data in set node, clean html with markdown node, convert file to csv and upload to an S3 bucket.

Something is messing the csv up (from S3 our data is loaded from a different service). In 1.44.1 it´s working perfectly in 1.45.0 something happens in the file conversion, which lead to a lot of rows being skipped on load.

In Spreadsheet tool like number, sheets, excel the file is aswell fine.

Describe the problem/error/question

What is the error message (if any)?

Please share your workflow

(Select the nodes on your canvas and use the keyboard shortcuts CMD+C/CTRL+C and CMD+V/CTRL+V to copy and paste the workflow.)

Share the output returned by the last node

Information on your n8n setup

  • **n8n version:1.45.0
  • **Database (default: SQLite):Postgres
  • n8n EXECUTIONS_PROCESS setting (default: own, main):
  • **Running n8n via (Docker, npm, n8n cloud, desktop app):npm
  • **Operating system:AWS Ami

Hi @Kool_Baudrillard, thanks for reaching out - at what point during the process are the rows skipped? When they are initially loaded from postgresql or when you are converting the file into csv format?

It would help if you could provide a simple version of the workflow that demonstrates the behavior you’re describing.

No,

the load by another system happens after the file is stored in S3.

But the csv generated in 1.44.1 and 1.45.0 seem to differ. The whole flow is fine and works in both versions as expected, the generated csv is the issue. I’ll clarify tomorrow with the dev team, if they see what happens on their side.

If you’d like to provide examples of the csvs before and after then that might be helpful as well to have available when looking at your workflow.

1 Like