Abstract

In recent years, privacy issues in the networking field are getting more important. In particular, there is a lively debate about how Internet Service Providers (ISPs) should collect and treat data coming from passive network measurements. This kind of information, such as flow records or HTTP logs, carries considerable knowledge from several points of view: traffic engineering, academic research, and web marketing can take advantage from passive network measurements on ISP customers. Nevertheless, in many cases collected measurements contain personal and confidential information about customers exposed to monitoring, thus raising several ethical issues. Modern web is very different from the one we experienced few years ago: web services converged to few protocols (i.e., HTTP and HTTPS) and a large share of traffic is encrypted. The aim of this work is to provide an insight about which information is still visible to ISPs, with particular attention to novel and emerging protocols, and to what extent it carries personal information. We illustrate that sensible information, such as website history, is still exposed to passive monitoring. We illustrate privacy and ethical issues deriving by the current situation and provide general guidelines and best practices to cope with the collection of network traffic measurements.