mda590
Can we get some distribution data ?
In My experience Averages are small part of the problem. Latency distribution or at least 99% percentile would be very helpful. Also information about packet lost is great as it really impacts distributed systems
I'd like to second Peter's suggestion, specifically a time series view of the latency. I ran a similar test a while back and found some surprising results when I viewed the data over time (published RTT data to ElasticSearch). For some AWS locations the latency had a step function specific hours of weekdays from other locations. I see "Graph showing latency over time" on your TODO list, hope we'll see that soon :-)
Percentiles have been added in the newly deployed version. I will look at adding packet loss information and add it to the enhancements list with the other items I am working on adding.
