I would use pg_copy if the data is not already in HDFS. Some parallelism in the copy might be possible, but chunking files in shell isn't particularly hard. Though if you do some benchmarking I would love to see the results!
-Mark On Sep 12, 2013, at 15:03, Matthieu Labour <[email protected]> wrote: > Hi > When exporting data from S3 to Postgres, I wonder when to use Sqoop over COPY? > Thank you for your help > -matt >
