GlideRecord Optimization for Service Portal Widgets

Quick Answer

This answer is correct because it addresses the two separate problems described in the scenario with two separate, targeted fixes rather than a single generic change. The slow loading is caused by the widget's server script running multiple synchronous GlideRecord queries against the same table, and replacing those with GlideAggregate where aggregated results are what is actually needed reduces the number of database round-trips the server has to make, cutting down the processing time before the widget can return data to the client. The blank page issue is a separate symptom coming from the client side: the client controller is not properly handling the asynchronous nature of data loading in Service Portal, so it tries to render the view before the server data has actually arrived. Using $scope.$on to listen for the point at which data becomes available ensures the controller waits for that data before rendering, which resolves the blank page rather than racing ahead with an empty or incomplete model. The key insight worth carrying forward is that portal performance problems and rendering problems often have distinct root causes even when they show up together, so diagnosing each symptom back to its actual source, inefficient server-side querying versus improper handling of asynchronous client-server data flow, and applying the specific fix for each is more effective than assuming one change will resolve both.