Concrete proof finally provided that the 2002 error has NOTHING is server side issue

[technole]

What they really need to do is enhance the launcher/login process to auto-rejoin the queue automatically if any 2002 shows up. If you get d/c'd, that spot should be saved longer as well. Among other QoL things that probably wouldn't take more than a day to add in the system.

Like who thought it was going to be acceptable to throw the error while you say 2k deep into a 4k queue and somehow didn't come back in-time to manually put your OTP back in and then lose your place. Awful design.

[MiaShino]

What they really need to do is enhance the launcher/login process to auto-rejoin the queue automatically if any 2002 shows up. If you get d/c'd, that spot should be saved longer as well. Among other QoL things that probably wouldn't take more than a day to add in the system.

Like who thought it was going to be acceptable to throw the error while you say 2k deep into a 4k queue and somehow didn't come back in-time to manually put your OTP back in and then lose your place. Awful design.

The same corporation that authorized setting up one single datacenter to service multiple countries both in the Americas and the EU then thought, this will end well.

Meanwhile nearly every other publisher including EA has the common sense to understand that a minimum of three geographical regions should be covered for NA. Western (Cali), Central (Texas), Eastern (Chicago). I try not to get sassy with these posts but some things just boggle the mind.

[Kes13a]

The same corporation that authorized setting up one single datacenter to service multiple countries both in the Americas and the EU then thought, this will end well.

Meanwhile nearly every other publisher including EA has the common sense to understand that a minimum of three geographical regions should be covered for NA. Western (Cali), Central (Texas), Eastern (Chicago). I try not to get sassy with these posts but some things just boggle the mind.

back in the day with dialup you would have a point. these days... not so much.

[MiaShino]

back in the day with dialup you would have a point. these days... not so much.

Huh? It is incredibly common for nodes across the Comcast i-bone to become oversaturated. This is.. basic networking stuffs here even for someone like me. It is even more important in modern era than dial up era to properly distribute data centers across large countries due to number of internet users constantly rising.

NTT nodes have been oversaturated for years resulting in still ongoing problems (90006). There is reason why most organizations have region based datacenters and this has been common practice for decades. When I self hosted I owned four virtual servers for good reason, spread out across the country.

what seems odd is that it obviously requires a condition to cause the 2002. what is the condition it requires?

Yoshi P himself;
"In order to prevent a major login server outage, FFXIV has a limit of "no more than 17,000 total players waiting to log in per logical data center". This cap is set uniformly regardless of how many Worlds exist in a logical data center.

・ The Issue

If the total number of players waiting in the logical data center exceeds 17,000, Error 2002 will be displayed and you will not be able to queue for login."

[technole]

The same corporation that authorized setting up one single datacenter to service multiple countries both in the Americas and the EU then thought, this will end well.

Meanwhile nearly every other publisher including EA has the common sense to understand that a minimum of three geographical regions should be covered for NA. Western (Cali), Central (Texas), Eastern (Chicago). I try not to get sassy with these posts but some things just boggle the mind.

It's not much an issue, EU used to actually be in the same data center as NA back when they were in Montreal with their previous provider before they went to Frankfurt.

Would help latency, but it's not so much about traffic as NTT has the infrastructure being a Tier 1 provider, they had been mostly fault-free since SE relocated to their Sacramento facility, it was said it takes 20 racks to run just one world. So having 500+ in one data center allows flexibility, saves on costs. If they use say the NTT Dallas, Texas hub, then if there was any problems routing to Sacramento, every player in those worlds would feel the faults. They would have to substantially outgrow the data center to consider the additional cost of a new location.

[MiaShino]

It's not much an issue, EU used to actually be in the same data center as NA back when they were in Montreal with their previous provider before they went to Frankfurt.

This is still issue. See;
https://forum.square-enix.com/ffxiv/...-REPLY-HERE%21
https://forum.square-enix.com/ffxiv/...NTT-Failure%29
Or simply browse the tech support section, you will see numerous 90006 related threads appearing especially during prime time. Me and my partner cannot play without a vpn anymore because of this.

Would help latency, but it's not so much about traffic as NTT has the infrastructure being a Tier 1 provider, they had been mostly fault-free since SE relocated to their Sacramento facility,

There have been many documented issues between Comcast/Charter i-bone and NTT interconnection. Many of these are result in oversaturated junction nodes loading into the NTT network.

it was said it takes 20 racks to run just one world. So having 500+ in one data center allows flexibility, saves on costs.

This is correct. It does save costs in short term.

If they use say the NTT Dallas, Texas hub, then if there was any problems routing to Sacramento, every player in those worlds would feel the faults. They would have to substantially outgrow the data center to consider the additional cost of a new location.

English is not the native language so I am having some trouble interpreting this, many apologies! There should never be routing for the player between regional datacenters. That is purpose of regional distribution. If Sacramento is experiencing issues then players connected to the Dallas data centers will not experience connectivity issues.

I cannot say for the certainty but it appears they have indeed long outgrown their Cali datacenter. I have no metrics though for this so who knows? Perhaps there is still room there for more racks or expansion of number of connections between Comcast and NTT.

[Kes13a]

What they really need to do is enhance the launcher/login process to auto-rejoin the queue automatically if any 2002 shows up. If you get d/c'd, that spot should be saved longer as well. Among other QoL things that probably wouldn't take more than a day to add in the system.

Like who thought it was going to be acceptable to throw the error while you say 2k deep into a 4k queue and somehow didn't come back in-time to manually put your OTP back in and then lose your place. Awful design.

what seems odd is that it obviously requires a condition to cause the 2002. what is the condition it requires? the process does work or no one would be able to log in period, so some factor is obviously causing the disconnect in the first place, and only happens sometimes, or no one would be able to play...period.

software either works or it doesnt, it doesnt decide "oh I dont like you, goodbye" there is a condition being met somewhere that is prompting some people to get repeated 2002 errors while other never or hardly see them. I have been in 3700 person queues, dont babysit and get in fine after my wait. others on here may get mostly through then hit, or get in only 10 or 15 deep before it happens.

there seems to be little reason behind what causes it, but it does not seem completely random if you believe the same people are having the issue consistently.

[Kotisimus]

the process does work or no one would be able to log in period

This is false, getting error 2002 doesn't prevent u from playing the game. It prevents u from logging into the game server IF u are not quick enough to just rejoin the queue.

or no one would be able to play...period.

Incorrect for the same reason above.

software either works or it doesnt

Again incorrect, software is not an ON/OFF switch, it can work and still have multiple bugs/issues. Please stop making assumptions based on ur own experience when u clearly got no idea.

[Kes13a]

This is false, getting error 2002 doesn't prevent u from playing the game. It prevents u from logging into the game server IF u are not quick enough to just rejoin the queue.

yes, but if it did not work AT ALL, then everyone would get 2002 and no one could log in. there are those that do not get this error. therefore, it is conditional. point out how that is "false"

Incorrect for the same reason above.

please feel free to post the entire part where I said, once in, the 2002 error would prevent you from playing?

Again incorrect, software is not an ON/OFF switch, it can work and still have multiple bugs/issues. Please stop making assumptions based on ur own experience when u clearly got no idea.

did you even read and understand what I said?

I said, there is some condition being met that prompts the 2002 error. if software runs into a certain condition, it will error. not everyone is seeing this error. so a condition is being met which drops the connection. this does not happen for everyone. it is possible, can the software work with a bug, it certainly can, but the programmers have a running joke about bugs and errors being caused by "unintended application" which means the user is doing something outside the regular recommended workflow. in this case, something in the connection is causing an unintended response.

no need to be antagonistic, I am in no way saying its not a code issue, there are issues with their code no doubt. just pointing out that it seems consistent for some, but not all. which leads me to suspect its more than just bad code. good to know your first forum post was so benign. I get it, you are frustrated, a lot of people are.

[Kotisimus]

yes, but if it did not work AT ALL, then everyone would get 2002 and no one could log in.

And i repeat again, getting 2002 DOES NOT prevent u from logging in. It just disconnects u from the queue and u can jump back in at THE SAME position if u are fast enough to do so. So u are absolutely wrong when u state that "no one could log in".

please feel free to post the entire part where I said, once in, the 2002 error would prevent you from playing?

You just said

then everyone would get 2002 and no one could log in.

Which is incorrect.

did you even read and understand what I said?

Yes, hence the suggestion to stop making assumptions when u clearly got no idea.

but the programmers have a running joke about bugs and errors being caused by "unintended application" which means the user is doing something outside the regular recommended workflow.

Me and my wife are playing over the same internet provider using exactly the same router from the same apartment. She will have error 2002 while i wont and in 1 hour i will have it and she wont. So i got no idea what the SE's "intended application" of their software is, but my wife and me where clearly using it as intended prior to EW since we were never getting error 2002 back then and suddenly we are NOT using it as intended after EW launch. Your logic is just flawed.