Re: sqlldr performance

If you mean 20 million records distributed among 30 tables, i.e. some rows to each table, but no row to more than one, and some field in the data being loaded determines which table gets that row, then you can run multiple sqlldrs specifying the table for that data in a

'when field_with_key = 'tablekey'
into table 'correct_table'

construct in the CTL file..
( see the utilities docs for more detail..) This will prevent more than one sqlldr writing to the same table.. ( If you need a more flexible solution, use a good ETL tool to distribute the data like Ascential Software's DataStage XE)

hth

nandagopalj_at_hotmail.com (Nan) wrote:

>Hello:
>
>I have a need to load about 20 million records in to 30 different tables.
>I am using sqlldr oracle utility to do this.
>
>My question is: To improve performance I plan on using multiple copies
>of sqlldr in a concurrent fashion. I am not sure of the impacts when
>multiple concurrent copies of sqlldr utility loads data into the same
>given table.
>I was told by the local DBA that it is not a very good approach.
>I am not fully convinced that it is a bad approach.
>I would appreciate your feedback on the same.
>
>Thanks,
>Nan.

-----= Posted via Newsfeeds.Com, Uncensored Usenet News =----- http://www.newsfeeds.com - The #1 Newsgroup Service in the World!  Check out our new Unlimited Server. No Download or Time Limits! -----== Over 80,000 Newsgroups - 19 Different Servers! ==----- Received on Wed Oct 17 2001 - 14:18:26 CDT