. Compression is critical since table size is quite tall, maybe 16-20 columns wide. My pipeline, as for now, is using Oracle GoldenGate on-prem, replicat on-prem as well in the same node, and build the … I was wondering what could be considered a best practice when it comes to extract incremental data from an Oracle transactional table running on-prem, and then pipe it to S3 as plain text …