Parallel job for loading an oracle table..suggestions needed
Posted: Tue Aug 26, 2008 8:33 pm
Hi,
I am writing a new parallel job for loading an oracle which was already written in server before. In server job it uses Sql Loader to load 300 million flat file in to a oracle table.
My first question is which one is the best approach to load, either to directly use sequential file to load the table or first make a dataset from flat file and then load it to oracle table.
My second question is right now server job using Sql Loader is being loaded in 2 and half hours and my aim is to reduce it atleast by half hour. parallel engine has two nodes.
So please suggest me a good approach which can cut down the load time.
Thanks in Advance
regards
Aakash
I am writing a new parallel job for loading an oracle which was already written in server before. In server job it uses Sql Loader to load 300 million flat file in to a oracle table.
My first question is which one is the best approach to load, either to directly use sequential file to load the table or first make a dataset from flat file and then load it to oracle table.
My second question is right now server job using Sql Loader is being loaded in 2 and half hours and my aim is to reduce it atleast by half hour. parallel engine has two nodes.
So please suggest me a good approach which can cut down the load time.
Thanks in Advance
regards
Aakash