11-20-2017 09:34 AM
I have a requirement to do full table loads for ~60 tables from an Oracle Database and I have a shell script that runs sqoop on each of those tables. But it takes a long time to load all those tables because some of them are huge, so I started tuning the sqoop job for each of them.
However, I stumbled upon this option "--fetch-size" and I have some questions related to it:
Does anyone know if it changes the "oracle.row.fetch.size" for the JDBC connection?
Is there a maximum limit for this parameter?
Does it impact the source DB or the Hadoop side resources?
Are there any guidelines about finding an optimum value for this parameter?
Thanks & Regards,