So I can't play and tell with EU/NA/JP friends in ARR...

[AarosLunos]

Didn't you just bash me for using Wikipedia? And no, that illustration doesn't apply.

Actually it does apply. SE very likely has their own ASN for BGP peering and route announcements. Assuming they build data centers in two other regions, they could perform BGP route announcements in such a way that reduces transit time.

Still doing lookups, found the follow ASNs so far:

AS17685 - PLAYONLINE SQUARE ENIX CO., LTD.
in -> as-block: AS17408 - AS18409

AS23784 - POLEXCHENGE SQUARE ENIX CO., LTD.
in -> as-block: AS23773 - AS23836

Going to take a look in those blocks might be they have some IP address ranges assigned for XIV specifically. I imagine SE, as a software business, requires a lot of bandwidth. MMOs even further expand that requirement. I imagine SE DOES have direct peering agreements with multiple Japanese carriers, throughout Europe and the Americas. Especially since they have offices in those locations.

[Starlord]

Actually it does apply. SE very likely has their own ASN for BGP peering and route announcements. Assuming they build data centers in two other regions, they could perform BGP route announcements in such a way that reduces transit time.

Yep I agree, It does relate to what Yoshida is actually trying to pull off, with Data centers being in different regions ( AS1, AS2,AS3 etc etc) and which every the "Customer" is the closest too, ( in that diagram being AS5), will be the Customers go to Data center. ( that is of course if I'm reading the diagram correctly )

[Mijin]

Actually it does apply. SE very likely has their own ASN for BGP peering and route announcements. Assuming they build data centers in two other regions, they could perform BGP route announcements in such a way that reduces transit time.

Well now your contradicting yourself. That's an illustration of interconnection and peering of autonomous systems, and you said the solution was a leased line. That illustration has literally nothing to do with this discussion.

You may also be interested to know that leased lines only exist in USA, UK, India, Hong Kong, and Italy.

[AarosLunos]

Well now your contradicting yourself. That's an illustration of interconnection and peering of autonomous systems and you said the solution was a leased line. That illustration has literally nothing to do with this discussion.

You may also be interested to know that leased lined only exist in USA, UK, India, Hong Kong, and Italy.

Incorrect leased lines can be built as virtual circuits attached to a physical interface. So they can be bundled over one physical medium and specific traffic shaping rules assigned. They are used all the time for over trans-oceanic fiber. I've configured plenty for domestic AND international clients.

It doesn't matter what the physical medium is since its piggy backed onto a physical interface at the router level. Its one of many ways to optimize routing. BGP route announcements (public and private) and leased lines can and often are used in conjunction. One common use is reducing the number of routers your traffic passes through while announcing your public IP address space to both ends of the leased line. BGP in itself doesn't optimize routing. That is done through peering agreements. BGP just announces the that addressable IPs are attached to a specific AS.

IF SE has peering agreements with specific providers it is possible they can change their BGP announcements to reduce latency without a leased line between the two other data centers. It just really depends because its situational.

I know that despite increasing evidence that I clearly know what I'm talking about, you won't believe me either way. If I wasn't bound by disclosure contracts, I would tell you specifically who I work for (HINT: Its a large backbone provider that is well known for pushing IPv6 roll out) But I'm not going to risk my career over a sniveling troll.

[Mijin]

Incorrect leased lines can built as virtual circuits attached to a physical circuit. They are used all the time for over trans-oceanic fiber. I've configured plenty for domestic AND international clients. It doesn't matter what the physical medium is since its piggy backed onto a physical interface. Its one of many ways to optimize routing. BGP route announcements (public and private) and leased lines can and often are used in conjunction.

No.

You're making your posts overly verbose and filling it with terms that don't even relate to hide the fact that you're just spouting random nonsense.

I'm going to make this really simple. Yes, hops effect your connection speed, but hops are something that is a result of your ISP, SE cannot control or improve them. While they effect your speed, the real problem with latency is coming from the DISTANCE of the the servers. They are far away, really far for most of us. The only way to solve this is to add new servers that are closer.

The fact is that Yoshidas statment, "The players will have a choice of selecting a server based in their region or a server overseas" doesn't have anything to do with peering, leased lines, or any other crazy things that don't relate to the issue. The thing that got everyone confused is the term "data center," and it's been explained many times now that a data center is just a building that houses servers. The only reason why Yoshida didn't elaborate further is because they were short on time.

Some of you feel strongly against this happening, and you want to believe is isn't. I can't do anything more to convince you, so I guess you'll just have to wait and see.

[AarosLunos]

No.

You're making your posts overly verbose and filling it with terms that don't even relate to hide the fact that you're just spouting random nonsense.

Actually I've been pretty detailed, considering I'm replying to a crappy troll on some forums. Apparently you don't know how ATM works or what it is. (In before Hurr durr Automated teller machine article on wikipedia)

Just noticed something else. Why is it all the troll posters tend to originate from the Sargatanas server? Seriously? Is it some weird conspiracy? Back when idiots where necroposting in 2 year old threads, or mass spamming topics. Mostly all Sargatanas players. Not that I'm generalizing everyone from that server. Just a weird observation.

[Mijin]

Actually I've been pretty detailed, considering I'm replying to a crappy troll on some forums. Apparently you don't know how ATM works or what it is. (In before Hurr durr Automated teller machine article on wikipedia)

"Hi, I can't win this argument, so you're a troll."

You are just unbelievable. I've seen a lot of people full of it on the internet but you take the cake. The simple fact is, what is causing the latency issue is the distance from our clients to the servers in Japan. The only way to fix that is to bring new servers to our region, and that's exactly word for word what Yoshida said he was going to do. Hops are a result of your ISP, and SE can't do anything about it. Everything you have said in this thread has been based on fiction, and a general lack of understanding of how the internet works.

[Tahldon]

"Hi, I can't win this argument, so you're a troll."

You are just unbelievable. I've seen a lot of people full of it on the internet but you take the cake. The simple fact is, what is causing the latency issue is the distance from our clients to the servers in Japan. The only way to fix that is to bring new servers to our region, and that's exactly word for word what Yoshida said he was going to do. Hops are a result of your ISP, and SE can't do anything about it. Everything you have said in this thread has been based on fiction, and a general lack of understanding of how the internet works.

a data center is just a building that houses servers.

I am bawling over here!

[AarosLunos]

"Hi, I can't win this argument, so you're a troll."

You are just unbelievable. I've seen a lot of people full of it on the internet but you take the cake. The simple fact is, what is causing the latency issue is the distance from our clients to the servers in Japan. The only way to fix that is to bring new servers to our region, and that's exactly word for word what Yoshida said he was going to do. Hops are a result of your ISP, and SE can't do anything about it. Everything you have said in this thread has been based on fiction, and a general lack of understanding of how the internet works.

I've won the argument and refuted all the misnomers and fallacies that you've presented with hard data and factual information. This will be my final reply to you. There are several others who have agreed with me thus far, and will continue to after my final reply. I guess you can never admit that you're wrong. All you insist on doing is calling into question my 12 years of experience in the industry. I guess you still didn't look up what ATM is yet. Its exactly what I described to you. While its an older technology its very much heavily in use and allows you to bind virtual circuits called PVCs to physical circuits Those PVCs can be assigned bandwidth limits and other traffic shaping rules.

Sounds like you need to also read up on the OSI model is because in any of your assertions you really fail to comprehend the differences between Layer 2 and Layer 3 routing. I mean the OSI model is the basic of basic. Its whats taught day one in entry Cisco certification courses. Sooo. Yea bye.