QoE has potential to be very helpful in troubleshooting application latency. Currently the "Time to First Byte" and "TCP Handshake" only show averages for the particular applications measured on specific monitored nodes/servers. There is no way to drill down and determine what these metrics look like from a specific ip address/host or subnet. All of our critical applications are hosted in our data center and accessed by remote branches across a WAN. There is no way to distinguish which hosts may be having poor "QoE" based on an average that includes both local and remote hosts. It would really be helpful to have this new feature include the functionality to drill down and identify specific hosts/ip addresses and their associated QoE stats.
↧