Where is the problem with this connection issue?

So lately I was having instant disconnects on specific occurrences, like if I turn on cut scenes on, I dc at the closing victory CS at the end of a DF (semi recent problem)
I also DC semi frequently teleporting to the lav bed sub ward where my house is.

I thought i trial WTfast to see if the issues continued and they are not.

If the server move is causing so much issues should SE be paying for these VPNs? because it feels like the move was more made to save money on servers, not upgrade them. I been at it with my ISP for over a month trying to see if something is wrong with the coax end of of the connection since a lot of speed downgrading happens there and may be causing a jitter issue. The lead tech wants to blame it on routing but how come I am like the only one suffering from this?

Does SE or my ISP have any influence on how data is routed? why is it so much better on a VPN? WHY DO I NEED TO USE A VPN while i LIVE in the US, connecting to a us server?????? Why can't SE code the game better not to spaz over a single packet loss? FFXI handles online connections better, why?

Also all the menus pop up faster when using wtfast.... also note I only have these issues with this specific game, not ffxi or wow.

VPN's bypass crappy routing on your ISP's side. It as nothing to do with SE. Also, you say the move was to save money on servers, but in order to do the move they had to buy all new servers, so that also makes no sense. the servers in Montreal were not physically moved.

https://i.gyazo.com/90c8e90340193dcb...c42eb3fe30.png

still happening with the vpn, guess I was just lucky before, the lav bed d/c never happened consistently soo...

I am not even sure if it happens just because where I am teleporting from. It always d/cs like this, as I am zoning in.

Quote:

---

Originally Posted by Snow_Princess

https://i.gyazo.com/90c8e90340193dcb...c42eb3fe30.png

still happening with the vpn, guess I was just lucky before, the lav bed d/c never happened consistently soo...

I am not even sure if it happens just because where I am teleporting from

---

I've been having frequent disconnects, but that is from everything, since my ISP is having issues atm. I haven't had issues with the game in a long time, except when they are being attacked remotely.

Quote:

---

Originally Posted by Valkyrie_Lenneth

I've been having frequent disconnects, but that is from everything, since my ISP is having issues atm. I haven't had issues with the game in a long time, except when they are being attacked remotely.

---

This happen to me quite a few time per day now.

So who is to blame with this? I still d/ced zoning in to lav beds with the vpn.

is my isp the one that determines all the routing? how does that work? is there more then just being in se's code or my isp?

Quote:

---

Originally Posted by Snow_Princess

So who is to blame with this? I still d/ced zoning in to lav beds with the vpn.

is my isp the one that determines all the routing? how does that work? is there more then just being in se's code or my isp?

---

Who is your ISP?

The ISP and the backbone providers are the ones who set the routes. Good ISP's route around bad network nodes to avoid interruptions, but most of them are lazy and don't do this.

Quote:

---

Originally Posted by Valkyrie_Lenneth

Who is your ISP?

The ISP and the backbone providers are the ones who set the routes. Good ISP's route around bad network nodes to avoid interruptions, but most of them are lazy and don't do this.

---

wide open west (WoW)

Network routing can be a little hard to explain, but yes, it's largly determined first by your ISP (typically called a last-mile ISP in the USA) which is then connected to a larger network. Simply, ISPs have ISPs too (Internet backbone providers), like Level 3, Cognet, Savvis, Verizon, NTT, etc. Content is typically cheapest if the ISP's ISP doesn't need to connect to something it doesn't own. ...it's sorta like having roaming charges for a cellphone.

What I'd recommend doing is running a traceroute to the servers you play on. Assuming that's mostly Balmung, like your user info says, you're on the Aether datacenter, which your client accesses at 204.2.229.9. Then, post in this thread in the FFXIV Technical Support section with the details.


Reporting Issues Resulting from the North American Data Center Relocation

To get the traceroute logs, you can run the following in your command prompt. It'll make a file on your desktop for you. It can also give you a better idea of where your connection might be failing. Or where it's going.

Quote:

---

tracert 204.2.229.9 > Desktop\ffxiv_balmung_tracert.txt

---

Most of the time, the core internet networks are gonna either be just an IP address or they'll have a regional-based name, so you can track the location that way. If you see a sudden spike in latency, that's where the issue is. If it's within the first/second hop, that's probably your home connection. If it's within the first few hops (you'll probably see your ISP's name here), then it could be their fault. If it's hit larger ISP ISP networks, you can try having your ISP's support file a support ticket for that ISP, but it's unlikely unless they're receiving a service outage or a massive amount of complaints. (Or they already have by the time you're affected). If it's happening on the last couple hopes, then it's on SE's side and likely affecting a large amount of users. ...and by then, we usually see a support topic on it.

does this tell you anything?
https://i.gyazo.com/1b36fa699b5d496f...1dec9ed20b.png

7 - 46 ms - 153 ms - 54 ms dynamic-75-76-35-10.knology.net [75.76.35.10]

17 - 113 ms 109 ms - 118 ms xe-0-1-0-1-1.r00.scrmca02.us.ce.gin.ntt.net [129.250.195.46]
18 - 197 ms 100 ms - 111 ms 204.2.229.2341
19 - 106 ms 106 ms - 102 ms 204.2.229.9

these hops seems like they are issues

Quote:

---

Originally Posted by Snow_Princess

does this tell you anything?

---

The reduced ping and jitter is good, but try the traceroute panel and see if it compares them. WTFast manually adjusts routing tables because it's a VPN. That's why it can make a better connection some of the time. (And why you have to generally pay for the service.)

You are still getting some packet loss though. Typically, games like FFXIV handle the connection over TCP so you won't actually get a loss of data. It'll just keep trying to be resent until it times out. (Or the client times out and forces a disconnect.) If you typically get a lot of lag or rubber-banding, that's from dropped/delayed packets.

But at 147ms average, the game should still typically play fine which makes me think it's more likely a packet loss issue at first glance. If your computer's connected over wifi, the signal may be weak or you may want to try a wired connection if possible.

edit: saw some of the traceroute stuff you added in.

See what's around hops 6-9. If it's consistently bad after 7, then that's probably one of the issues. 17-19 is ok because that could just be network hardware trying to stay "hidden." A lot of backend devices will reject being pinged to reduce the chance of DDoS attacks.

Quote:

---

Originally Posted by reiichi

The reduced ping and jitter is good, but try the traceroute panel and see if it compares them. WTFast manually adjusts routing tables because it's a VPN. That's why it can make a better connection some of the time. (And why you have to generally pay for the service.)
You are still getting some packet loss though. Typically, games like FFXIV handle the connection over TCP so you won't actually get a loss of data. It'll just keep trying to be resent until it times out. (Or the client times out and forces a disconnect.) If you typically get a lot of lag or rubber-banding, that's from dropped/delayed packets.

But at 147ms average, the game should still typically play fine which makes me think it's more likely a packet loss issue at first glance. If your computer's connected over wifi, the signal may be weak or you may want to try a wired connection if possible.

---

It is not wifi, so what am i doing now? I edited in some hop things though the cmd prompt but a bit paranoid to copy the full thing to the public.

I just do not understand having these issues when in the same country!

Quote:

---

Originally Posted by Snow_Princess

It is not wifi, so what am i doing now? I edited in some hop things though the cmd prompt but a bit paranoid to copy the full thing to the public.

I just do not understand having these issues when in the same country!

---

It could possibly be your ISP. It's tough to say with the information provided. Hops 6-9 would give me a better picture since those are definitely outside your local area, but would appear to still be going through your ISP. Once things hit NTT, that's SE's provider for the current North American datacenters.

Quote:

---

Originally Posted by reiichi

It could possibly be your ISP. It's tough to say with the information provided. Hops 6-9 would give me a better picture since those are definitely outside your local area, but would appear to still be going through your ISP. Once things hit NTT, that's SE's provider for thecurrent North American datacenters.

---

6-9
6 25 ms 28 ms 39 ms static-76-73-191-224.knology.net [76.73.191.224]
7 46 ms 153 ms 54 ms dynamic-75-76-35-10.knology.net [75.76.35.10]
8 52 ms 45 ms 48 ms static-75-76-84-231.knology.net [75.76.84.231]
9 44 ms 48 ms 58 ms dynamic-75-76-35-105.knology.net [75.76.35.105]


NTT:
10 56 ms 68 ms 59 ms ae-53.r08.nycmny01.us.bb.gin.ntt.net [129.250.194.81]
11 44 ms 58 ms 48 ms ae-3.r24.nycmny01.us.bb.gin.ntt.net [129.250.5.61]
12 84 ms 82 ms 97 ms ae-2.r20.sttlwa01.us.bb.gin.ntt.net [129.250.4.13]
13 90 ms 89 ms 88 ms ae-0.r21.sttlwa01.us.bb.gin.ntt.net [129.250.2.54]
14 104 ms 99 ms * ae-3.r23.snjsca04.us.bb.gin.ntt.net [129.250.3.124]
15 104 ms 109 ms 119 ms ae-45.r01.snjsca04.us.bb.gin.ntt.net [129.250.3.175]
16 115 ms 109 ms 107 ms ae-1.r00.scrmca02.us.bb.gin.ntt.net [129.250.7.31]
17 113 ms 109 ms 118 ms xe-0-1-0-1-1.r00.scrmca02.us.ce.gin.ntt.net [129.250.195.46]
18 197 ms 100 ms 111 ms 204.2.229.234
19 106 ms 106 ms 102 ms 204.2.229.9

Perfect. I can give you infodump now.

Your connection is being routed to New York City first, which then heads over to Seattle, and that's where you're packing on the most latency. (Hop 11 to 12). Depending on your ISP, the NYC NTT backbone is probably the closest/cheapest to go to for them, since NTT is mostly East/West Coast and a couple big cities. Central USA and the Midwest has basically nothing. (You can see NTT's network map of their backbone connections here: NTT Network Map

So depending on where you live, your connection might be heading East before it can go West, towards the datacenters. The longer the distance and the more hops it takes, the higher the latency and higher chance of a dropped packet. If you have other ISPs that service your area, they could be better/worse depending on their networks and the backbones they use. But everything will still point to the easiest NTT device, and then go from there.

Using that information, you could check with your ISP if there's a better route to NTT's network for your area. Maybe they're taking an unoptimized path or they can adjust some of their routes which might improve your connection to the game.

Quote:

---

Originally Posted by reiichi

Perfect. I can give you infodump now.

Your connection is being routed to New York City first, which then heads over to Seattle, and that's where you're packing on the most latency. (Hop 11 to 12). Depending on your ISP, the NYC NTT backbone is probably the closest/cheapest to go to for them, since NTT is mostly East/West Coast anda couple big cities. Central USA and the Midwest has basically nothing. (You can see NTT's network map of their backbone connections here: NTT Network Map

So depending on where you live, your connection might be heading East before it can go West, towards the datacenters. The longer the distance and the more hops it takes, the higher the latency and higher chance of a dropped packet. If you have other ISPs that service your area, they could be better/worse depending on their networks and the backbones they use. But everything will still point to the easiest NTT device, and then go from there.

Using that information, you could check with your ISP if there's a better route to NTT's network for your area. Maybe they're taking an unoptimized path or they can adjust some of their routes which might improve your connection to the game.

---

That is so BS! why can't it go to san jose or settle directly? why is it going to NY first? who controls that? I was just talking to the lead tech in my area by my ISP and hes basically saying " I do not know we might not even control that"

Quote:

---

Originally Posted by Snow_Princess

That is so BS! why can't it go to san jose or settle directly? why is it going to NY first? who controls that? I was just talking to the lead tech in my area by my ISP and hes basically saying " I do not know we might not even control that"

---

A support tech likely has no control over it, but knology.net belongs to Wide Open West, so all the routing is done internally. So at least you're not being handed off to another provider in the middle. Since NTT is largely just backbone in the USA, it's easy to get a network map of their stuff. Other backbone providers also have them sometimes. (There was this really cool one I had at work for a while, but it was discontinued.)

I can't find a very good network map for WoW, but their business pages had this? WoW Network Map for Business

If it's the same case for consumer stuff, all the network is basically built out to service a particular set of states, so they're gonna take the shortest (read as: least amount of hops on their own network) to get to the nearest NTT network, which in your case, appears to be the New York location. I have to assume you're likely more geographically located to it than Chicago, which is where the next nearest NTT network would be.

When using WTFast, your connection probably ignores a lot of that and just heads to the nearest "big" backbone provider and gets a nicer route, which is why things like ping, jitter, and packet loss go down.

If you've got local friends who also play FFXIV, see what provider they have and if they have the same connection issues. Compare traceroutes and so on. Unfortunately, it's unlikely that your ISP is gonna build out expensive infrastructure unless they start expanded to additional markets. ...we can't even get the big-name providers to build out better infrastructure in a timely manner.

Quote:

---

Originally Posted by reiichi

A support tech likely has no control over it, but knology.net belongs to Wide Open West, so all the routing is done internally. So at least you're not being handed off to another provider in the middle. Since NTT is largely just backbone in the USA, it's easy to get a network map of their stuff. Other backbone providers also have them sometimes. (There was this really cool one I had at work for a while, but it was discontinued.)

I can't find a very good network map for WoW, but their business pages had this? WoW Network Map for Business

If it's the same case for consumer stuff, all the network is basically built out to service a particular set of states, so they're gonna take the shortest (read as: least amount of hops on their own network) to get to the nearest NTT network, which in your case, appears to be the New York location. I have to assume you're likely more geographically located to it than Chicago, which is where the next nearest NTT network would be.

When using WTFast, your connection probably ignores a lot of that and just heads to the nearest "big" backbone provider and gets a nicer route, which is why things like ping, jitter, and packet loss go down.

If you've got local friends who also play FFXIV, see what provider they have and if they have the same connection issues. Compare traceroutes and so on. Unfortunately, it's unlikely that your ISP is gonna build out expensive infrastructure unless they start expanded to additional markets. ...we can't even get the big-name providers to build out better infrastructure in a timely manner.

---

I am IN CHICAGO! as in MAIN PROPER, not suburb that thinks they are part of our city, the MAIN CITY. I thought it was obvious what I said about going to san jose or settle directly and why i am now pissed off about this. Ty for identifying the problem though, now i got more ammo to my isp mrgrgrgrgr

Quote:

---

Originally Posted by Snow_Princess

I am IN CHICAGO! as in MAIN PROPER, not suburb that thinks they are part of our city, the MAIN CITY. I thought it was obvious what I said about going to san jose or settle directly and why i am now pissed off about this. Ty for identifying the problem though, now i got more ammo to my isp mrgrgrgrgr

---

Good luck! Hopefully they can tell you why you're being routed like that. ...or figure out why they're routing like that because it could be a mistake. I don't know enough about gritty networking details or how routing tables are set up, but it could be the case that their hardware's unaware of the NTT Chicago hub, or it was down at their last check, or something else and it was determined that the "better" prioritized route was to NYC. It's probably adjustable although it might require someone with more authority than a customer service tech or a field agent to handle.

The DCs happen when;

You change zones.
Someone else zones out/leaves the PT (right at the exit, not simple 1 of those, but both at the same time)
Enemy attacks that treat "moving you" as a zone teleport.
Being revived. (Same as above. Its a same zone teleport, treated as zoning)

Changing zones is the easiest to notice, as u do it more, but the others do happen.
2nd most common is when trying to get ur loot at the end. Someone leaves, causing u to DC, and if the chest was opened after u fully DCed, u will come back to no loot to roll on. (sucks in savage PuGs)

Also, its a bit weird, as I can have 2 versions of the game running on my PC, as I use keybord, and my friend uses controller.

One of us will DC, while the other doesnt. (same PC, same internet)
so its not as simple as our connection truely goes out.

plus it doesnt happen with any other game/MMO
just FFXIV. (happened before the server moves too.)

have 2 roomates who play in other rooms, who wont DC when 1 person does.

we know its a true dc when we all dc at the same time.

is this still going to NY?

dynamic-76-73-171-85.knology.net 3 0.05 5699 5696 10 35 799 39 1314 65 29.55
user-24-214-131-140.knology.net 4 0.02 5699 5698 10 28 694 22 524 61 24.74
76-73-164-101.knology.net 5 0.05 5699 5696 11 29 743 28 544 63 25.93
static-76-73-191-224.knology.net 6 0.00 5699 5699 12 31 660 29 825 57 24.81
dynamic-75-76-35-10.knology.net 7 0.02 5699 5698 12 30 683 29 766 58 24.36
dynamic-75-76-35-6.knology.net 8 0.09 5699 5694 12 31 743 24 900 56 26.35
13335.ch.equinix.com 9 0.09 5699 5694 11 31 680 55 794 59 26.78
Target WTFast.com 10 0.07 5699 5695 12 32 719 39 1077 60 27.36

Really like to know how wtfast is doing the routing. It will help me know what that is to help me sort for a new isp.

called ISP today, phoned with different people in the course of 5 hours. This is what I got:
"We do not control the routing but we can send a tech to your house to look at it"
"We always send it to New York first, that is how it has always been and never changed it"
Me: not acceptable anymore in this day and age
yeah im dropping them soon. tried of the lag from the server move SB gave, hey on the bright side help me realize my ISP is garbage.

using vpn:
2 12 ms 11 ms 11 ms unknown.puregig.net [173.245.202.1]
3 13 ms 13 ms 15 ms 38.142.65.161
4 12 ms 12 ms 14 ms ae-11.r06.chcgil09.us.bb.gin.ntt.net [129.250.9.121]
5 13 ms 13 ms 12 ms ae-1.r21.chcgil09.us.bb.gin.ntt.net [129.250.3.7]
6 90 ms 82 ms 96 ms ae-5.r22.snjsca04.us.bb.gin.ntt.net [129.250.5.17]
7 95 ms 96 ms 87 ms ae-19.r01.snjsca04.us.bb.gin.ntt.net [129.250.3.27]
8 87 ms 87 ms 88 ms ae-1.r00.scrmca02.us.bb.gin.ntt.net [129.250.7.31]
9 93 ms 95 ms 98 ms xe-0-1-0-1-1.r00.scrmca02.us.ce.gin.ntt.net [129.250.195.46]
10 87 ms 146 ms 88 ms 204.2.229.234
11 87 ms 94 ms 86 ms 204.2.229.9

Trace complete.

also this seems to give me the best performance I can find atm:
3 32 ms 36 ms 21 ms 4.35.180.213
4 30 ms 24 ms 22 ms ae-3-3503.edge4.Chicago3.Level3.net [4.69.203.138]
5 22 ms 22 ms 31 ms 4.68.111.70
6 83 ms 73 ms 78 ms ae-21.r06.chcgil09.us.bb.gin.ntt.net [129.250.3.158]
7 25 ms 25 ms 26 ms ae-1.r21.chcgil09.us.bb.gin.ntt.net [129.250.3.7]
8 76 ms 70 ms 71 ms ae-5.r22.snjsca04.us.bb.gin.ntt.net [129.250.5.17]
9 75 ms 74 ms 75 ms ae-40.r02.snjsca04.us.bb.gin.ntt.net [129.250.3.121]
10 74 ms 75 ms 72 ms ae-2.r00.scrmca02.us.bb.gin.ntt.net [129.250.7.33]
11 72 ms 74 ms 73 ms xe-0-1-0-1-1.r00.scrmca02.us.ce.gin.ntt.net [129.250.195.46]
12 73 ms 74 ms 76 ms 204.2.229.234
13 78 ms 71 ms 72 ms 204.2.229.9

how does this fair?

Those both seem to be pretty good. Disconnects are more from delayed/dropped packets, so unless you're also getting bad lag, a more stable connection is typically the better thing to go after. Granted, less hops and a more direct route usually improve that because it's less chances for something to go wrong.

If your connection to the game seems good with those connections, I'd stick with them. A ping under 100ms is usually seen as extremely good for most MMOs/multiplayer games. (Still depends on how often polling occurs, but in games like FFXIV, that's typically every 200-300ms).

This is the one that seems to be performing the best:
Packets: Sent = 1025, Received = 1017, Lost = 8 (0% loss),
Minimum = 69ms, Maximum = 215ms, Average = 77ms
Packets: Sent = 14109, Received = 13977, Lost = 132 (0% loss), Minimum = 69ms, Maximum = 418ms, Average = 77ms

3 29 ms 15 ms 24 ms te-9-5.csr2.Chi3.Servernap.net [66.252.0.178]
4 21 ms 27 ms 14 ms ae-12.er1.Chi2.Servernap.net [66.252.0.70]
5 24 ms * 25 ms Pccw-br04-Chc01.er1.Chi2.Servernap.net [66.252.3.134]
6 48 ms 28 ms 23 ms ae-19.a01.chcgil09.us.bb.gin.ntt.net [129.250.8.137]
7 90 ms 80 ms * ae-2.r05.chcgil09.us.bb.gin.ntt.net [129.250.5.95]
8 35 ms 17 ms 23 ms ae-0.r21.chcgil09.us.bb.gin.ntt.net [129.250.2.205]
9 77 ms 72 ms * ae-5.r22.snjsca04.us.bb.gin.ntt.net [129.250.5.17]
10 75 ms 79 ms 80 ms ae-40.r02.snjsca04.us.bb.gin.ntt.net [129.250.3.121]
11 86 ms 78 ms * ae-2.r00.scrmca02.us.bb.gin.ntt.net [129.250.7.33]
12 * 79 ms 78 ms xe-0-1-0-1-1.r00.scrmca02.us.ce.gin.ntt.net [129.250.195.46]
13 * 74 ms 75 ms 204.2.229.234
14 71 ms 71 ms 73 ms 204.2.229.9

given each hop and ms time, what ya think?
I doono how you feel about what i am gonna say next but this is how i feel from testing things:

I feel a real VPN is getting me better performance then wtfast. It is skipping knology nodes and seems to have lower threashoalds in pings (76-72 never for wtfast) and seems to have lower spike values.

Go for it.

Sometimes other VPNs work better than WTFast, or other "Gaming" VPNs like Pingzapper or MUDFISH. Use what works best for you!

https://i.gyazo.com/77fb3edf0133ef7e...86c6e83dc7.png
no d/cs so far like the experience I got under my ISP, I just found it weird it kept happening at the same occurrence (teleporting to lav beds and d/c during the final CS in DFs)
I just do not understand how it was my ISP when it keeps repeating for the same situations.

Quote:

---

Originally Posted by reiichi

Good luck! Hopefully they can tell you why you're being routed like that. ...or figure out why they're routing like that because it could be a mistake. I don't know enough about gritty networking details or how routing tables are set up, but it could be the case that their hardware's unaware of the NTT Chicago hub, or it was down at their last check, or something else and it was determined that the "better" prioritized route was to NYC. It's probably adjustable although it might require someone with more authority than a customer service tech or a field agent to handle.

---

They are blaming it on SE:
http://forum.square-enix.com/ffxiv/t...-extra-routing.

Reaching destination = success I guess, this is more proof for me SE really needs to test in less then optimal conditions for their coding. The game is way too picky what happens on a single packet loss or how the game is effected by 70-80 ms vs 130-150 ms

Basically I am blaming both, SE for the coding, expecting perfect net, and my ISP not understanding what quality is.

whats this one doing (testing a diff isp)

1 <1 ms <1 ms <1 ms homeportal ----------
2 20 ms 20 ms 20 ms ------------.lightspeed.cicril.sbcglobal.net ------------
3 21 ms 21 ms 21 ms ---------------
4 * * * Request timed out.
5 26 ms 23 ms 23 ms 12.83.43.57
6 25 ms * 22 ms cgcil402igs.ip.att.net [12.122.133.161]
7 22 ms 22 ms 22 ms 192.205.32.194
8 80 ms 80 ms 78 ms ae-11.r05.chcgil09.us.bb.gin.ntt.net [129.250.3.144]
9 21 ms 23 ms 50 ms ae-0.r21.chcgil09.us.bb.gin.ntt.net [129.250.2.205]
10 75 ms 74 ms 75 ms ae-5.r22.snjsca04.us.bb.gin.ntt.net [129.250.5.17]
11 79 ms 80 ms 79 ms ae-40.r02.snjsca04.us.bb.gin.ntt.net [129.250.3.121]
12 79 ms 80 ms 79 ms ae-2.r00.scrmca02.us.bb.gin.ntt.net [129.250.7.33]
13 80 ms 79 ms 79 ms xe-0-1-0-1-1.r00.scrmca02.us.ce.gin.ntt.net [129.250.195.46]
14 79 ms * 85 ms 204.2.229.234
15 86 ms 84 ms 86 ms 204.2.229.9