What’s Wrong with Network Positioning and Where Do We Go From Here?

Network positioning systems are essential to good performance in distributed systems that scale to millions of end hosts. Evaluating performance in this environment is particularly challenging. This paper addresses this issue through an empirical study of two alternative classes of network positioning services based on a dataset gathered from more than 43K IP addresses probing over 8M other IPs worldwide. We use more than 1.4 billion network measurements to show that network positioning exhibits noticeably worse performance than previously reported studies. To explain this result, we identify several key properties of this environment that call into question fundamental assumptions driving network positioning research. We close by suggesting a new direction for network positioning that draws from our observations.