Neither of those tests is to the ffxiv servers. You need to be looking at your specific path to Ormuco's 199.91.189.0/24 subnet that the game's servers are hosted on in Montreal Canada. And your bandwidth is not an issue with this game, generally speaking. About the only time you need more than a DSL-Lite level of bandwidth is for patching, heavily congested areas, and initial loading. And even then, you would be more then fine with a 15Mb bandwidth plan. What is of greater concern is latency---it needs to be consistent, and it needs to be low.
Any test to anywhere other than the network hosting the game is only going to show you issues on your localized networks at best. Looks like you are in or otherwise coming through the Florida area---you will NOT have a direct line to Montreal from there. You will be handed off to a third party network. Typically, we see a lot of issues either around that hand off, or between there and Ormuco (SE's ISP in Montreal). So the route taken specifically for your game client is what needs to be looked at.
Oh, and even though that trace is not to the game service in Montreal...it does show a lot of jitter in a few spots, which can cause problems. Even in your more local segments you have a swing just over 3x your average response times. Further down, you see the trend continue to 3x and 4x the average time (not to mention it is an even bigger jump from the best times). And that TWC hop with the intermittent dropped packets is a bad sign. Even though we are talking forwarding in the route, it is a clear indication that that hop has reached a high enough utilization point that it is ignoring lower priority traffic trying to preserve throughput. In other words--that hop is trying to prevent congestive failure---which is not good.
The game relies on timely, orderly delivery of the packets. If packets get too far out of order/delayed you're risking a retransmit which can get progressively worse when there are lag spikes kicking in. Overly congested nodes present a problem for this game. The packets aren't encrypted (like with a VPN), and aren't otherwise flagged to put them on a higher priority list. This makes them vulnerable to getting delayed during periods of high congestion....which is the problem more often than not. Since they cleaned up my routing earlier this year, EVERY time there has been any instability that was not announced to be an issue server-side, it was because either my local devices were flaking out (modem or router), or there was an overly congested segment along my route--which Time Warner would get me routed around and service would improve.


Reply With Quote

