Problem: my “lowest latency” WAN is superficially low due to poor routing decisions inside their network. For example, pinging a random website (ESPN):
min/average/max/stddev
WAN1: 31.103/32.035/33.459/0.695 ms (the “lowest latency” WAN)
WAN2: 13.833/24.788/33.609/6.216 ms
I’d be nice if the “lowest latency” algorithm could be tuned to get a more complete picture of how the WAN connects to the internet. Or (if I can dream), it’d be nice if the “fastest response” algorithm could feed its implicit latency measurements into the “lowest latency” algorithm.