past 3 nights in less that 5 mind after login 90002

[Domini]

so i play both ps4 and PC I been getting sudden 90002 midday once, and mainly evenings cannot do anything but disconnected every few min after logging in
It needs to be addressed as it wasn't doing this earlier week.

I've changed Ethernet cables,used wireless ,used wired, cycled router, used different router ,modem, contacted isp all good.


I Am On Hyperion server, So others get it cities dungeons quarry-mill. Please square fix this game isn't playable in this frequent disconnecting state

or least look into the 90002 disconnects. its not normal to dc 5 mins after logging I counted 8 times tonight in little time.

[Raist]

Need to check your route to the server you play on. There are known issues with routing in both North America and abroad--something SE is virtually powerless to address, and it falls to our ISP's to work with their partnered ISP's who provide peering/transit to other people's networks. It's a problem with the Internet Service Provider industry in general, and NOT server specific--so you need to flush that out first, as it is the most common issue found across the board.

Get the IP FFXIV is using by loading Resource Monitor after you have been fully logged into the game for a minute or so (resmon from the search/run box, look in the network or TCP sections). Then run tracert to that address from a command prompt (cmd from the search/run box, then as an example, tracert 199.91.189.31 <ENTER>) to check for anything that looks out of sorts. Copy/paste the report into an email to your ISP's Tier3 support for a proper review (right click anywhere in the CMD window for a pop-up menu to select all text, then press <ENTER> to copy all the highlighted text to clipboard for pasting into notepad or an email, or support ticket form). Here is a post about the process put up by Blizzard for running this to diagnose routing/connectivity with their servers:

https://us.battle.net/support/en/art...g-a-traceroute

[Frost_Trunks]

Yeah I beg to differ. I use WTFast and I still get 90k'd several times a night. It's rather annoying. My wife does not use WTFast, and she get's booted just as often.

[Raist]

Yeah I beg to differ. I use WTFast and I still get 90k'd several times a night. It's rather annoying. My wife does not use WTFast, and she get's booted just as often.

Using a VPN won't necessarily prevent you from hitting the same (or another) congested or otherwise troubled network segment on the way to the same destination. While there is a very long and varied list of potential paths to get around the US, there is a considerably shorter list for getting around Canada--and an extremely short list for getting specifically to SE's data center in Montreal. Last time I looked them up, Ormuco had only 5 peers (Level3, Cogent/PsiNet, TATA, TiNet/SPA, and ALTER.NET/VerizonBusiness/UUnet)--and that fifth one was added only recently. Haven't done a full dig on that Verizon peering arrangement yet, but of the other 4, NONE of them had more than 3 paths into Montreal at least semi-directly from a US city--some had a fourth path, but it was a potentially long-range sling-shot around Seattle/Vancouver and a handful of hops to Toronto/Ottowa to get there (very bad if you are on the East Coast)--but none had more than 3 that were a straight shot from a US city to Montreal. The bulk of all the paths were going through the North East...If I remember correctly, there were only 3 that weren't through the North East and took multiple hops to get to Toronto first, and even through the North East it was only like 7 or 8 total--including a route that slung up around Nova Scotia or something.

The point is... regardless of where you are coming from, if ANY of Ormuco's peers are having issues, you have a pretty strong chance of hitting that troubled path. Let's say you are assigned Cogent, who shows only three direct paths into Montreal from a US city. Lets say they are having issues with the Albany-Montreal leg, and the route you are assigned puts you on Cogent and is sending you to the DC/Ashburn exchange. Out of there, you basically have three options from that point--pathing to Albany, Cleveland, or Boston--two of those are pretty much going to shoot you to Albany. ONLY the Clevelend path provides you an alternate path besides Albany, and that is a handful of extra hops to get around to Ottowa and Toronto.

Basically, you have a 3 in 4 chance of hitting the troubled path between Albany and Montreal if you are coming through Washington via Cogent in that scenario. So what needs to happen is to alter your path so that you either come through a different region, or across a different ISP's routes to avoid that bad corridor into Montreal. Keep in mind that the options are very limited for the end of our routes--like around 10 or so possible corridors to Ormuco. So at BEST, you have around a 1 in 10 chance of hitting a troubled corridor if just ONE is having issues during peak time. Now... consider if all 5 ISP's are having issues in one specific region (which there is far more potential than many may realize--it's not uncommon for at least 3 of them to have issues all at once in somewhere like New York, New Jersey, amongst a long list of other hotbeds for activity). Your chances for hitting a floundering corridor just ramped up big time if you have 3, 4 or 5 out of the 10 potential paths exhibiting problems.

Ultimately... the point is, you can't just fix your routing by simply connecting via a VPN. Sure, their admins may be more aggressive about monitoring routes and you have a considerably better chance of avoiding congested nodes---but it is far from perfect implementation. That is why you see options to manually select alternate tunnels. It may seem strange to hear it, but sometimes you may find you fair better by tunneling to a site further away than the one you automatically get assigned. So, if you connect with a VPN and it is the same (or even worse)---go into the configuration and try a different location.

You can do a "quick" test before logging into the game to see if an alternate tunnel may prove more stable by running tracerts. Make note of who's names show up along your route between you and Ormuco (they will be one of the last two hops prior to the ones starting with 199). You may also notice the DNS names give some hints as to where the hops are located. Not all of them will be as obvious, but most will use at least a 3 letter notation for the city--some even use airport abreviations in some locations. Then pick another tunnel site and run another trace to see if those ISP names or at least the locations change for those last few hops before Ormuco. If you see big changes, you may have just skirted around the congestion--log into the game and test the connection out.

[T2teddy]

I do not understand 95% of what you're saying Raist, but be that as it may, my friends and I live in all 4 corners of NA and the 4 of us are getting DCed right, left, and center simultaneously. And there's no way people in Alaska, Nova Scotia, BC, Florida and Texas are all using the same route.

[Raist]

I do not understand 95% of what you're saying Raist, but be that as it may, my friends and I live in all 4 corners of NA and the 4 of us are getting DCed right, left, and center simultaneously. And there's no way people in Alaska, Nova Scotia, BC, Florida and Texas are all using the same route.

But you can all be routed through the same region that is having trouble. Look at the Cogent map I linked to in that earlier post. That is the map for one of the 5 ISP's that your ISP connects to in order to get you to Ormuco, who is SE's ISP. Look at the connections around Montreal, then read through the scenario again. That scenario can play out equally with any of the 5 ISP's that Ormuco peers with. If the big four that route mostly through the NY/NJ/Maine corridors (Level3, TATA, Cogent, Verizon) are all having issues with the same region (be that weather phenomena, excessive use because of something like Superbowl or March Madness streaming)--it can affect LARGE groups of people at once, regardless of where they live. It is because they are being routed through over saturated corridors. It causes big blotches on the outage maps for lots of the last mile ISP's, because there are regional congestion issues impacting connectivity across the routes.

Here's a post I made a couple months back with links to articles, apps that track usage, as well as a few outage maps that update periodically throughout the day..it may help some of you guys gain a better understanding of just what is really in play with our connectivity woes:

http://forum.square-enix.com/ffxiv/t...=1#post2765690

Just in case some won't be bothered to read up on it, I'm going to try to re-paste in the three maps from that post and the source pages for them here as well:

https://downdetector.com/status/comcast-xfinity ----- https://downdetector.com/status/att ------ https://downdetector.com/status/time-warner-cable



Those maps are archived images of services tracked at the downdetector site. The images will get updated periodically throughout the day, so may change when loaded again later. But if you look closely, you can see patterns forming in specific regions. It's all tied into the infrastructure issues discussed in those blog posts.