In our application which uses a Facebook Graph API, in the Application-Level Rate Limiting dashboard, we can track three parameters: Call Count Usage Rate, Total Time Usage Rate, and Total CPU Time Usage. I've noticed that our application very often exceeds a Total Time Usage Rate. Unfortunately, I couldn't find how it is calculated a Total Time Usage available for our application calls. Does anyone know how is computed?
I've tried to find this info in official documentation, but there is only explained how Call Count Usage is calculated.
Facebook documentation: https://developers.facebook.com/docs/graph-api/overview/rate-limiting/