A company uses a custom application with a table named 'project_task' that contains several reference fields pointing to large tables (e.g., 'User', 'Department'). Users report that the 'project_task' form takes over 10 seconds to load, primarily due to the time required to fetch options for these reference fields. The administrator wants to improve the form loading performance without removing any fields or changing the user workflow. The instance is on a standard license and performance optimization is a priority. Which approach should the administrator take?
This loads reference options asynchronously, allowing the form to render quickly while options populate in the background.
Why this answer
Enabling 'Use Ajax on Reference' on the dictionary records causes the reference fields to load their options asynchronously via AJAX after the form renders, rather than during the initial page load. This significantly reduces the initial load time because the large reference tables (e.g., 'User', 'Department') are not queried synchronously when the form is first opened.
Exam trap
The trap here is that candidates often think a reference qualifier (Option C) will improve performance by reducing the dataset, but they overlook that the query still executes synchronously during page load, whereas 'Use Ajax on Reference' changes the loading mechanism itself to be asynchronous.
How to eliminate wrong answers
Option A is wrong because converting reference fields to choice fields with a predefined list would require manual maintenance of static lists, which is impractical for large, dynamic tables like 'User' and 'Department', and would break referential integrity. Option B is wrong because writing a client script with GlideAjax to load reference options asynchronously is a custom approach that duplicates built-in functionality, adds unnecessary complexity, and is not the standard or recommended optimization for this scenario. Option C is wrong because adding a reference qualifier limits which records are displayed but does not change the synchronous loading behavior; the query still runs during page load, so it would not address the 10-second delay caused by fetching all options.