I benchmarked Spring Batch vs. a simple JobRunr setup for a 10M row ETL job. Here's the code and results.