Has anyone experienced issues with load balancing on the PI Web API? I have some applications that are heavily using the PI Web API and have noticed that when many users are accessing the applications, large batch requests tend to come back all at the same time across all users. It feels to me that the PI Web API is not running with a multi-threaded/asynchronous approach.
As a simple example, I could have two pages that make batch requests with normal response times of:
Page A - 10 seconds
Page B - 5 seconds
If User1 launches Page A immediately before User2 launches Page B, then both requests take about 10 seconds to complete. However, if User2 launches Page B before User1 launches Page A, then the response times are about what is expected.
Are there some configuration settings I can apply to improve these performance issues?