I expose a public API that is backed by a SLSB. As these are publicly available in an environment where multiple tenants are accessing the application, I would like to throttle API usage for each user.
My thought is if I could access a threadlocal or context setting for SLSBs in the pool, I could determine how much concurrent usage is happening in my system.
1) Is there a better approach to accomplishing this?
2) If this technique is as good as any, how can I access the thread pool and/or SLSB pool during a request in order to make this work?