We recently did a round of inventory counting, where we created a set of count reports segmented by a particular categorization; we invoked INVENTORY_PART_IN_STOCK_API.Create_Count_Report to do this for each of these. Unfortunately some of these categories, even though we are only interested in “top-level” handling units, contain a lot of items when you include all the levels of nesting. One category, containing over 25,000 handling units across all of these levels but with <1000 top-level items, took over 16 hours to complete. We’ve had a category with an amount in this neighborhood (more like 20,000) that took closer to 2 hours to run, which was still problematic enough that we came up with special-case subcategories.

My boss suggested modifying this method (or its implementation submethods) to skip certain checks or otherwise run much more efficiently, and before trying that I wanted to know if this is just innate to the count report process, if there is another setting I should tweak (there are a lot of defaults passed to Create_Count_Report), if there is a preferred way of creating a count report, etc.