I have pyspark code that writes to an s3 bucket like below:
df.write.mode('overwrite').parquet([S3_BUCKET_PATH])
I am testing writing to the bucket via the bucket's access point instead. The AWS documentation has an example writing to the access point using the CLI like below:
aws s3api put-object --bucket arn:aws:s3:us-west-2:123456789012:accesspoint/prod --key my-image.jpg --body my-image.jpg
I have tried doing it like this:
df.write.mode('overwrite').parquet("arn:aws:s3:us-west-2:123456789012:accesspoint/prod")
However, I get this error:
Relative path in absolute URI
Is it possible to write to an S3 access point using pyspark?