Message 11–20 sur 25
par
ColtsWalker4 ans plus tôt
I am seeing more entity warping (creatures rapidly changing positions) on white and yellow bar recovery. Recovery is rapid, however, more frequent entity warping is occurring. Root problem continues to be your rack space source. There is a bottleneck in the edge router to the country where the game server is located. Ping times >200 makes game play problematic. The best ping times I ever see are around 150ms. They get up around 300 and I close AD and go do something else. In short: The game client does recover from bars now much better, however, more creature warping is observed.
par
Firefly4 ans plus tôt
I can only investigate and fix problems originating from my server and its data center. I've dug very deep into packet captures, network traces, pings and even the kernel TCP/IP stack, but all signs point to problems outside of my control. I currently can't fix anyone's local network router being pressured by other users. I currently can't fix overloaded Wi-Fi access points. I currently can't fix sub-ocean, intercontinental fiber optic that is dropping packets due to increased demand probably stemming from COVID. I can't fix oversubscribed ISPs not being able to meet peak traffic demands during weekends and evenings. After analyzing packet captures for about twenty hours I've yet to find one where my server and data center are the problem. Angeldust is normally 100% playable globally, but right now there seem to be extraordinary circumstances causing a global rise in network traffic. Pain points have emerged where traffic seems to be dropped. For an interactive game that's killing, but I'm not sure I can fix this without having to roll out my own global network. @Hummm has helped diagnose and capture many problem instances and that's been very helpful. Still, there are no signs that anything within my control is wrong. I can't fix anything besides my code and own services. I'm absolutely willing to fix anything within my control though, but someone needs to step up and provide proof before I can do something. 300ms ping is absolutely no problem, I've tested up until 800ms ping and the game is still very much playable. The important factors are ping (latency) variance and packet loss. Normally (before COVID) this wasn't a problem, but it seems like many more people are homebound now and overloading key parts of the Internet. I hate to put the blame onto others and/or outside my control, so please give me reputable leads and packet captures or network traces to investigate. (Word of warning: like with the griefer claim solution discussion I'm pretty sure we'll eventually go down to the brain-though-sharing level where I go and point out how statements are no proof, no solution and no lead for investigation. Be advised that it's not a lack of willingness, but that I do require very specific and detailed input to be able to fix anything.)
Message 11–20 sur 25
Rob12
Hummm
