I sincerely don't think SE's netcode can handle this.
Printable View
I sincerely don't think SE's netcode can handle this.
Are...are you going to make a new topic every time they change slides?
One of Japan's biggest game publishers and IP holders having their online services incorporated into AWS and their growing monopoly on internet infrastructure and multiplayer video games is a great thing. As an Amazon shareholder, I'm laughing all the way to the bank. Don't forget to pay your student loans everybody, the moratorium ends in a few weeks. :)
Im starting to notice a very obvious Pattern that Im only now catching on to......
Dont know if Im upset about it, or impressed it stealthed me for this long.
Technical/spicy take:
I've always had a theory that SE's servers run on rather outdated hardware and solving performance issues/migrating to newer architecture etc is problematic (Eg: look at how long EU servers had horrible rubber banding in duties for back in SB).
If that is indeed correct, running the servers in the cloud (Aka basically running on AWS or Azure etc) will give SE far more flexibility to scale the capabilities of the DC up or down as required and most importantly, they should be able to scale it virtually overnight.
As long as the entry point isn't problematic, aka ping, I think it's a neat idea and I'm glad to see them try it.
Oh? What's that?
They have improved the hardware sometimes with infrastructure maintenances.
Yes, that is the point. The reason they decided against the cloud is because it's slower. However, it also allows them to respond to congestion such as what we get at an expansion launch. So that is why they are making it an option but not making it replace the physical servers. They believe physical is faster.Quote:
If that is indeed correct, running the servers in the cloud (Aka basically running on AWS or Azure etc) will give SE far more flexibility to scale the capabilities of the DC up or down as required and most importantly, they should be able to scale it virtually overnight.
Outside of Koros troll doing Koros troll things… they’re doing a testing phase first and can work it out from there. Honestly I think it’s a good thing to consider if it gives them a backup or contingency for the future. I also wonder if this means they can do more down the line infrastructure wise. Take the physical off to upgrade and transfer to cloud while doing the work for instance or go to cloud only I dno