oh Tom, not sure how to thank you, you explained in such a simple and clear way. I was banging my head to understand what am I trying here, your explanation made me think otherwise. A big Thank you from the bottom of my heart. You just made my day. Sorry to bug you,but few clarifications. " So the physical files in that directory should be in all lower case with an extension of .sas7bdat." :--> I just need to specify the table name without ".sas7bdat" extension right? 1. If i need to create multiple hive tables, can i do that ? something like data myhive.myhivetable ;
set mysas.mysasdataset ; data myhive1.myhivetable1 ;
set mysas.mysasdataset; 2. Based on your experience, which is better approach? approach being followed now OR converting SAS to .csv and then loading it to hive? 3. SAS is on SERVER A and Hadoop is on SERVER B. Can i call the script from server B(Hadoop) rather than server A? If csv is a better approach, I would call it from sever B and create an external hive table and point the location from where I am creating csv. 4. Finally, I may need to load SAS dataset in 500GB size, is there any way to speed it up? I have access to only UNIX box and I am not sure what all options are possible. Once again, Thanks Tom for taking your time and replying in detail. Good day Sir..
... View more