compute real packet loss and latency variance from PCAPs
Right now packet loss and latency variance information displayed in charts comes from test scenario definition, but in theory it might differ from real values - depending on number of packets and servers in the test.
It would be better to compute real packet loss and latency variance from PCAPs and display these values.
It would work as a) sanity check b) more precise data visualization, especially if variance is huge (e.g. some scenarios from networks in China).