Has anyone benchmarked ETL EMR?

Hi Snowplowers,

I was wondering whether anyone has benchmarked ETL-EMR before based on number of thrift records and AWS Core/Tasks instances types?

My jobs on 500 Thirft files are running on 3 r3large + 10 Tasks r3larges and Step2 has been running for over an hour. That seems too slow?

Regards
SS

1 Like