Something wrong here Anet

Same connection, same path

JQ/GOM 6,780 bytes sent (122ms to server), return traffic received 58,530 bytes (1050ms). No person within 20k units.

Now here is SoS 6,590k bytes sent (125ms to server), return 90,974k bytes (435ms). No person within 20k units.

Same rotation, same build, same npc's etc and in same order. Graphics options are all minimum. Connection speed is 245/20mbps.

EBG: mounts have decreased the time taken to reach a given unit area of population conflict and increased the amount of calculations required before return. What is alarming is the differing rtt between the servers with all things remaining equal. That should not be taking place. Further, I am assuming the server priorities have also been set to regional timeslots as I'm Aussie and my rtt during SEA is a good 50% less than if I play in a NA timeslot with the former having far more population in combat during the oversight period.

Incidentally, VPN does not change the outcome as the time taken to calculate seems to be the root cause, although I must admit the Amazon return is also disgraceful and am in full agreement with our Nation Broadband Network (NBN) techs in saying that the Amazon server priorities need a review.

Is Aussie broadband any good. Heck no.. garbage stuff in most states and that is being downright forgiving. However.. the latency problems do not stem from tcp path up, but rather, in the array and return path. Why not consider the Blizzard solution which was to place a regional server in Australia or SE Asia. Amazon has the infrastructure and servers in ie Australia, Hawaii, Singapore etc and the exchange rate would be of benefit in the cost of the service fees.

When you say regional server you mean a separate region, right?

Otherwise you seem to have spent quite alot of time arguing infrastructure without considering the infrastructure.