View Single Post
Lt. Commander
Join Date: Dec 2007
Posts: 120
# 37
01-31-2010, 05:42 PM
Originally Posted by Grtishord
I can see Queues working but if this is true then why all the rampant disconnects and failure to launch instances? If the queues were to prevent this then why is it still occuring? I agree that they have a shot (slim but there) of fixing these problems before launch. It will take downtime, more servers, this is the headstart for those of us who pre-bought. What is going to happen Tuesday at 10AM PST when the thousands more who didn;t want to pre-buy come in like barbarians crashing the gate?
Probably all that is needed is more headroom on the servers (i.e., more processing power, more robust components); they clearly can't handle the present traffic. I'm also not clear on this "shard" set-up. Is it multiple servers linked to one central hub in some kind of distributed I/O arrangement, or is it more complex than that? Is this a more economical ("cheaper") or more efficient way to do it (or both)? Does that mean the log-in server is the choke-point? Or is it just that the aggregate hardware array is under-powered for the demand and needs to be expanded? Seems like software tweaks from now until launch would be too little too late to fix this.