>I have a mortgage application that needs to load 200,000 transactions from an ASCII file at the end of every month. This process is taking 3-4 hours on the clients PC (350MHz, 64MB RAM). The client wants to be able to load up to 1 million. I am looking for the most efficient way of loading, processing and saving these transactions.
>
>Some ideas:
>1. Split the load file between multiple computers
>2. Sort the load file and the table by loan number and use scan/skip
>3. Try to make code more efficient (difficult)

I have an app that reads through and parses a 40 -60 MB text file (about 500,000 to 750,000 lines) on a daily basis on an IBM Thinkpad, P3 with 128 MB RAM. It only takes a matter of a few minutes to bring in the file, parse the data, and e-mail individual pieces of the report to various credit unions. I'm not sure of the nature of your data or what you do with it when you bring it in, but it shouldn't take that long. You may want to look at making the code more efficient.

>4. Get the client to use a faster PC.

Faster PC is always a good thing.

>5. What else can I do?