I just used the import feature. I don’t know the exact steps, but importing a dBase 5 table with 256 fields and literally millions of rows takes easily an hour. I was told by a Microsoft developer that the reason for this is SQL Server is mapping many more field types to every record than dBase 5 supports. For example, in dBase, it is either a number (int) or float. But in SQL Server, it could be a short, int, long, double or float.

Eventually, I abandoned SQL Server for MySQL. It can import a dBase 5 table of that size in 10-15 minutes, which is much much faster than SQL Server 2000. And, being the FOSS advocate that I am, it follows my beliefs and principles, where proprietary software such as SQL Server 2000, does not. MySQL also does everything I need, plus much much more. I love it!

Using a DTS package direct copy is in my experience the fastest way to batch load your database and it can process 100′s of thoughsands of records rapidly.

Of course this does depend on your table structure e.g. A triggers, indexes , etc will slow it down.

Have you tried Oracle (the free version… unless you’re gonna splash out)?

You should try SqlLite – it is small and VERY fast. Not as robust as MySql, but it sounds like it may be better for what you are trying to do. Give it a try.