In 0.111.0, we had a simple workflow that called an API endpoint to get data, transform that data and insert it in a Postgres table. The workflow looked like this:
The workflow worked absolutely fined until we updated our version to 0.119.0.
Starting from that point, the “Execute workflow” throwed a formatting data “NaN” error on the last Postgres node.
BUT if we opened that node and click on “Execute node” everything worked perfectly fine and the data were inserted into the table correctly.
We thought maybe something was weird in our data formatting function and tried changing the code in various ways, the error persisted. We also looked directly into the workflow file and didn’t notice anything abnormal.
We then updated to 0.123.0 and now even the “Execute node” was throwing the “NaN” error, just like “Execute workflow”. At least it was consistent, but we still couldn’t figure why the code was suddenly not working.
In desperation we decided to rollback to 0.111.0 (not changing any piece of the workflow/code) and everything was absolutely fine again. The “Execute workflow” and “Execute node” didn’t throw any error.
We do not have any idea of what could cause this on the most recent versions.
We do not think that could be the issue.
If it can help, we are going to try to replicate the problem on an other instance with public data during the next days, and share the workflow here.
To keep you updated: We tried to upgrade our Node version to 14+ but the error persisted on the recent n8n versions. Unfortunately that wasn’t the problem.
I don’t think I understand what you mean by “applying the changes to the node”.
After upgrading Node.js and n8n to the last versions, I tried to execute the whole workflow again, and had the error.
I get that the nodes return values from all the queries now, but each Postgres node contains only one query.
I have the same error but only to float Type.
After read query I casted field to float by parseFloat(fieldname) but I have some null value to the field too.
I update my item function from: