We have some copy jobs that use s3distcp which runs on EMR cluster deployed on EKS pod. These jobs copy data from S3 to HDFS and vice versa. We are unable to run this after getting the Hadoop cluster kerberized. Tried multiple options of passing keytab and running kinit but it's failing either at mapper or reducer end based on it's a pull or push request. Any way to do this without kerberizing EMR cluster?
Tried all options of passing keytab and running kinit but s3distcp still fails as it's not getting token to connect to Kerberized Hadoop cluster
