5 Foot Step wrote on Sep 24
th, 2018 at 12:27pm:
Correlation =/= causation. Take a moment to consider that the entire game population is now less than a single shard population during the peak around MotU. Now realize that prime raid time is also prime internet time in general. If you get unplayable lag at prime time right now (I don't), then you are still going to lag even if you are the only one logged in to the server.
A lot has changed for DDO on both the codebase and the hardware since 2012 so using performance metrics and information from that era as the measuring stick in 2018 wouldn't seem to be very valid.
How many entirely new in-game systems have they added since then? How many updates and changes have they made to existing code from that time? Ditto for the hardware. It was physical hardware back then and it's apparently a VM environment now. Remember that 3-month long lagfest after the datacenter migration?
And he's right to some extent. Currently on Ghallanda at least, it becomes pretty obvious when SSG hasn't restarted the world servers because the lag just keeps getting worse and worse, especially during primetime hours. It is bad enough at times that it most definitely causes otherwise preventable deaths, wipes, etc. which interfere with missions and raids.
DDO lags like a mofo sometimes while everything else running through my inet connection works just fine. ICMP, traceroutes, bandwidth tests, etc. I've heard rumors that Ghallanda lags more for *reasons*, but since SSG hates their customers and refuses to provide any sort of transparency, I can only assume they're full of shit based on their track record and that it is due to G-land having one of the larger primetime populations combined with a trainwreck of a codebase. Small inefficiencies in code become much larger ones as you increase the usage load on them.