How to optimise the job with huge source tables?

Hi All,

I have 2 source tables having lots of records(227354374 and 46526852) that I need to load into my target structure in BW.

I am using Merge transform to load the data into my target.

I'm a bit concerned about the performance issue that this huge number of records can create when I start the load.

Is there any performance tuning techniques that I can use or a way to split the number of records into smaller chunks so as to avoid any performance glitch?

Regards,

Ankit