An experimental evaluation of Distributed Rate Limiting for cloud computing applications

Pricing and load distribution in a cloud are difficult problems. In order to recover the construction costs of the data center, the owner must endeavour to have high utilization while maintaining performance levels for the users. Distributed Rate Limiting (DRL) is a mechanism that can be used to attack this problem. A mathematical framework and the convergence and stability properties of two DRL algorithms were examined in previous work. In this paper we build upon that work by evaluating the performance of these algorithms on a testbed against the results of a simulation carried out in previous work. We also evaluate the algorithms resilience to failure and propose the good-neighbours enhancement to improve its performance.