Data issue when pushing data from s3 to redshift

Question

Hey all, I currently have a data file in s3 which i am trying to output to redshift through glue visual studio. My data successfully got sent to redshift tables, but the amount field has changed the values to scientific type 
For example- 500303030 became 5.558543197E7	in redshift. How can i solve this issue? the column is varchar type in redshift, and string type in the glue node before which it gets copied to redshift tables. Thanks!

Accepted Answer

Hello,

Thanks for reaching out. For this scenario, a workaround is to add an intermediate node called "Change Schema" in Glue studio, which converts the input column into "string" before writing data to Redshift.

So the data flow is as below:

Data source, S3 => Change schema => Data target, Redshift.

The generated code for "Change schema" looks like below, for sample only:

# Script generated for node Change Schema
    ChangeSchema_node1702948869240 = ApplyMapping.apply(
        frame=AmazonS3_node1702946282328,
        mappings=[
            ("invoiceno", "string", "invoiceno", "string"),
            ("amount", "bingint", "amount", "string"),
        ],
        transformation_ctx="ChangeSchema_node1702948869240",
    )

Hope it helps.

Answer

You may need to leverage CONVERT function to convert from scientific number format to float: https://docs.aws.amazon.com/redshift/latest/dg/r_CONVERT_function.html

Data issue when pushing data from s3 to redshift

関連するコンテンツ