We have been toying around with Availability Group Auto Seeding across replicas. So far so good we have been successful and even implementing TDE encryption in the mix across 4 Azure replicas. One issue we faced was the testing VM's did not have much horse power and we ran out of worker treads. As a convenience we decided to build a quick worker thread alert so we know when we are near the limitations. This can have other useful purposes. In testing this seems to work pretty well. Hope some others can make use of the code base. I grabbed a quick query from Dr. Google and went at it, 10 min had a working alert.