I don't have much insight into spark but I've been using Dataflow/beam for ETL. Been a pretty good experience. follows the style of spinning up compute to proce

I don't have much insight into spark but I've been using Dataflow/beam for ETL. Been a pretty good experience. follows the style of spinning up compute to process as needed then shutdown.