Measuring and Injecting Latency in Web Apps

In mobile computing and other contexts, latency is a more critical resource than bandwidth. Interactive performance relies on low latency response to the user. The question of how much performance is necessary (and when) is still open. These sources of latency manifest themselves in different ways and with differing severities. When designing systems for a Web 2.0 framework, it is important to have a good means for measuring existing latencies as well as simulating new latencies, throughout all parts of a system. This report discusses a web app latency analysis framework with respect to a set of representative tasks. These tasks include: web usage, navigation with tabs, navigation between many open windows, and significant copying and pasting. This research was supported by the National Science Foundation (NSF) under grant number CNS-0833882. Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF or Carnegie Mellon University.