Anatomy of a Throttler, part 2

This text delves into the design considerations for implementing a database throttler with a comparison of singular vs distributed throttler deployments. It explores the pros and cons of both approaches, including high availability concerns in singular designs and the introduction of multiple layers of collection polling intervals in distributed designs. The text also discusses various strategies to reduce the impact of the throttler on the system, such as adjusting the probe frequency based on request patterns and hibernating metric generation during periods of low activity.