I have a traffic manager in Azure, with routing method as "performance".
Currently it has two endpoints, westus-a
and eastus-a
, each endpoints serves 1k/min request. An hour ago I added another endpoint westus-b
to this traffic manager. After waiting for ah hour later, the
traffic to westus-a
only decreased a little portion, and westus-b
only serves 200/min requests.
According to this document:
If your profile contains multiple endpoints in the same Azure region, then Traffic Manager distributes traffic evenly across the available endpoints in that region.
If this is the case, the westus-a
and westus-b
should both serve around 500/min request.
Also according to another document, the "performance" routing method relies on the internet latency table to choose the least endpoints.
Both westus-a
and westus-b
are deployed in West US region, there should not be so much difference in terms of latency among these two endpoints.