API Strategy & Practice Conference 2018

Throttling is the threshold for limiting the number of requests to a component. This threshold is important wherever an invocation chain passes through several distributed components. If throttling is not configured correctly, the infrastructure is at risk of accidental or malicious overload.

For example, a call passes from an API consumer through various layers of the architecture before it reaches the system of record where the server responds to the request. By taking time to correctly design the throttling implementation, you can dramatically reduce the risk of overload.
The session will cover throttling concepts and key considerations for five throttling scenarios. It describes throttling in the context of architecture. For example, using throttling to protect a service host or to limit a user to the agreed service-level agreement (SLA).