Finding Multiple Requests In Large Data Sets

[Teutonic_Night]

I have a solution that helps keep track of file deliveries. We delivery relatively large numbers of files (so far, up to about 90k per month), and some of these deliveries are of the same file to multiple destinations. I need to create a list of all new deliveries in a given month, and all previous deliveries that have been invoiced.

I already have a method to get a list of unique filenames, but I now need to search the "delivery" table for all instances of each unique filename delivered in a given month.

For example, the table "deliveries" contains 1,000,000 records. In August we made 25000 deliveries, of which 15,000 were unique files. I need to search all 1,000,000 for every delivery for each of the 15,000 unique files that were delivered in August.

Right now, I have a method that works but is awfully slow. It essentially performs 15,000 finds, with 14,999 expanding the initial found set. I was wondering if anyone had any suggestions for a method to perform a find such as this that would be a bit faster. I'm already using layouts with no fields on them and using the "set field" script command to set up my search criteria. I'm not searching on any calculated fields.

This is on a hosted database, hosted by FileMaker Server Advanced 11, with FileMaker Pro 9 clients used to access it.

[comment]

This is rather confusing. Why do you need to search the "delivery" table for all instances of each unique filename delivered in a given month - instead of finding all deliveries in a given month?