Toward Next-Generation Distributed Rate-limiters

Abstract

The cloud services made accessible over different network paths and served by multiple backend servers demand distributed rate-limiting solutions. Emerging technologies such as 5G/6G, IoT, and Edge cloud are changing the network landscape dynamically and demand more sophisticated rate-limiting policies. However, the current designs of Distributed rate limiters (DRLs) demand to tradeoff the overheads (communication, computation, state/memory) with accuracy in terms of configuring the rate-limiters with parameters like the number of connections, bandwidth, etc. Hence, we propose accurate and low-overhead containerized distributed rate-limiting functions managed by hierarchical software-defined networking controllers.