Setting ACL in S3 objects written by an AWS Glue Job

Question

I came across the following issue:

I run a Glue job from account A to write into an S3 bucket in account B. This meant that the owner of the object is account A and I couldn't do anything with the objects from account B.

Is there a way to tell the Glue job to apply an ACL with full control for the bucket owner?

Accepted Answer

You can use Hadoop java configurations in a spark job to update the S3 canned acls. I haven't tested this with Glue Dynamic Frames, but it works for native Spark DataFrames.

import sys
    from pyspark import SparkConf
    from pyspark.context import SparkContext
    from pyspark.sql.functions import *
    from pyspark.sql import SQLContext, Row

sc = SparkContext()
    sc._jsc.hadoopConfiguration().set("fs.s3.canned.acl", "BucketOwnerFullControl")

Setting ACL in S3 objects written by an AWS Glue Job

Relevant content