These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

Issues, Workarounds & Localization

 
  • Topic is locked indefinitely.
Previous page123Next page
 

The socket was closed

First post First post
Author
Previous Topic Next Topic
Gektor Davo
Imperial Shipment
Amarr Empire
#21 - 2016-10-15 15:01:57 UTC
I tried many things to do)
Turning on DMZ with my comp ip, turning on ipv6 protocol, clean cache, reinstalling client, cable/wifi both connections, closing launcher when after get in game etc. etc. etc....Nothing works for me. I still fear to kill NPC's, cause i allready lost 1 machariel like that and i still cant create a new character.
Thanks, CCP. From Russia with love *
Shin Kudo Udan
Caldari Independent Navy Reserve
Curatores Veritatis Alliance
#22 - 2016-10-15 15:07:51 UTC
Same problem here since yesterday morning (I'm from Argentina). I don't know how to read the network diagnostics test results, but I think there's a problem there...

Quote:
Starting ping for client.eveonline.com

Haciendo ping a client.eveonline.com [87.237.34.202] con 32 bytes de datos:
Tiempo de espera agotado para esta solicitud. (Request timed out)
Tiempo de espera agotado para esta solicitud.
Tiempo de espera agotado para esta solicitud.
Tiempo de espera agotado para esta solicitud.

Estad¡sticas de ping para 87.237.34.202:
Paquetes: enviados = 4, recibidos = 0, perdidos = 4
(100% perdidos),
Verdi Ohvant
Aliastra
Gallente Federation
#23 - 2016-10-15 15:08:27 UTC
I am having the same issue, though ignoring Cert errors did not fix it for me.

Only way I can log in is to use a cellular hotspot for my ISP.

Would love to hear some input from CCP on this.
xstonex
School of Applied Knowledge
Caldari State
#24 - 2016-10-15 15:17:26 UTC
Same problem can't connect since yesterday downtime tried everything suggested above can we please get ccp to look at this.
Flakout
Caldari Provisions
Caldari State
#25 - 2016-10-15 16:23:28 UTC
Yeah not sure been having same issue since last DT. Hopefully its an easy fix only thing that might work is i completely log my accounts off (From launcher). Then constantly keeped logging in til it worked but seem like i have to do this every time i log an account in.
Bedja
Science and Trade Institute
Caldari State
#26 - 2016-10-15 16:29:15 UTC
Kattara wrote:
Ok, so it seems that the new launcher has some nice diagnostics built-in. So the issue appears to be here:

---
K:\Program Files (x86)\CCP\EVE\Shared\tq\code.ccp\requests\packages\urllib3\util\ssl_.pyj:318: SNIMissingWarning: An HTTPS request has been made, but the SNI (Subject Name Indication) extension to TLS is not available on this platform. This may cause the server to present an incorrect TLS certificate, which can cause validation failures. You can upgrade to a newer version of Python to solve this. For more information, see https://urllib3.readthedocs.org/en/latest/security.html#snimissingwarning.
.
.
.
K:\Program Files (x86)\CCP\EVE\Shared\tq\code.ccp\requests\packages\urllib3\util\ssl_.pyj:122: InsecurePlatformWarning: A true SSLContext object is not available. This prevents urllib3 from configuring SSL appropriately and may cause certain SSL connections to fail. You can upgrade to a newer version of Python to solve this. For more information, see https://urllib3.readthedocs.org/en/latest/security.html#insecureplatformwarning.
---



Fixed the problem by deleting everything in shared cache folder still getting errors in log. But client is stable
Gizzie Haslack
4249003
#27 - 2016-10-15 16:59:07 UTC
Bedja wrote:
Kattara wrote:
Ok, so it seems that the new launcher has some nice diagnostics built-in. So the issue appears to be here:

---
K:\Program Files (x86)\CCP\EVE\Shared\tq\code.ccp\requests\packages\urllib3\util\ssl_.pyj:318: SNIMissingWarning: An HTTPS request has been made, but the SNI (Subject Name Indication) extension to TLS is not available on this platform. This may cause the server to present an incorrect TLS certificate, which can cause validation failures. You can upgrade to a newer version of Python to solve this. For more information, see https://urllib3.readthedocs.org/en/latest/security.html#snimissingwarning.
.
.
.
K:\Program Files (x86)\CCP\EVE\Shared\tq\code.ccp\requests\packages\urllib3\util\ssl_.pyj:122: InsecurePlatformWarning: A true SSLContext object is not available. This prevents urllib3 from configuring SSL appropriately and may cause certain SSL connections to fail. You can upgrade to a newer version of Python to solve this. For more information, see https://urllib3.readthedocs.org/en/latest/security.html#insecureplatformwarning.
---



Fixed the problem by deleting everything in shared cache folder still getting errors in log. But client is stable



I cleared the cache and still get dumped off of the server after 1 hour of gameplay.

Logging in is fine. It's just staying logged on for more than an hour where I have the problem :/
ISD Max Trix
ISD Community Communications Liaisons
ISD Alliance
#28 - 2016-10-15 21:38:25 UTC
CCP Is aware that there is a network issue and is working on possible issues.

ISD Max Trix

Lieutenant

Community Communication Liaisons (CCLs)

Interstellar Services Department

I do not respond to EVE mails about forum moderation.
Gizzie Haslack
4249003
#29 - 2016-10-15 21:39:21 UTC
ISD Max Trix wrote:
CCP Is aware that there is a network issue and is working on possible issues.



Thankies :)
Fer'isam K'ahn
SAS Veterinarians
#30 - 2016-10-15 22:36:37 UTC
Hy, I also have some issues. I know someone said CCP was made aware, but maybe some of mine will help find the issue.

- Since DT today I get disconnected exactly every hour after login.
- Always the first instance on one PC gets the socket closed... all other run without interruption (always the same two drop out of four).
- Rebooting PCs, Router, changing Network, switching WiFi to Cable and vice versa has no effect.
- The disconnecting instance can be detected during login, because it gets a grainy screen (you have to zoom in to see that the black is grainy speckled with grey). Never seen that before, nor do the others share that screen. Weird thing.

Hope that helps (and get fixed). I was fleet booster today and the one with the Leadership skills and implants.. dropped every hour and fleet comp got busted 8(

o7
IChooseYou
Tradors'R'us
IChooseYou Alliance
#31 - 2016-10-16 03:06:25 UTC
I just got this error after multiple failed logins: The specified proxy or server node (1272308) is dead

Buying all pilots at the best prices out there!
Gizzie Haslack
4249003
#32 - 2016-10-16 12:15:54 UTC
IChooseYou wrote:
I just got this error after multiple failed logins: The specified proxy or server node (1272308) is dead



Sanshas. Or Drifters.

What? :p


I'm pretty sure that it's a Datacentre issue. One of the centre's switches is on the way out. They work fecking hard, so do drop from fatigue.
Bedja
Science and Trade Institute
Caldari State
#33 - 2016-10-16 14:07:15 UTC  |  Edited by: Bedja
ISD Max Trix wrote:
CCP Is aware that there is a network issue and is working on possible issues.


Would be nice that CCP would inform us, insted of telling us to try VPN access and so on.

Quote:


https://support.eveonline.com/hc/en-us/articles/201827699--The-Socket-was-closed-message

"The Socket was closed" message

Updated Today at 14:00


The error "The socket was closed" is the result of the TCP protocol closing the connection as it deems the connection unreliable. This usually is the result of the protocol having to resend the same packet several times in a row without success and unfortunately can happen very quickly for a variety of reasons, reaching from interference on a wireless network or a squeezed network cable to a router experiencing issues of some sort outside of the jurisdiction of ours or your ISPs circle of influence.

Any issues within our own hardware or close to our end that could cause this would be caught quickly by our monitoring and affect any player whose connection goes over that part of our networking hardware at the time and thus affect a large player base at the same time, and we usually also post an according note on our status twitter @EVE_status when we become of aware of issues that may cause disconnects.

If everything on your ISPs end is fine as well, the root cause for those issues likely lies on other hops along the route. If you have access to a VPN service that allows you to connect via other servers in the world, it can help checking whether or not those connection losses still appear when going through this different route.



Got it working yesterday by clearing out the cache but today it's back to normal 15min then it drops
Gizzie Haslack
4249003
#34 - 2016-10-16 14:56:54 UTC  |  Edited by: Gizzie Haslack
Bedja wrote:
ISD Max Trix wrote:
CCP Is aware that there is a network issue and is working on possible issues.


Would be nice that CCP would inform us, insted of telling us to try VPN access and so on.



Indeed. Slow & clunky cheap VPN's, or my Fibre at full speed... Pretty obvious I'll be wanting the latter.

Which worked fine until 2 days ago.



As for the " it's in the cloud " claim; don't fall for it. The major data pipes linking continents are huge. Most likely their data-centre is having issues and has not told them.

As in, CCP are being blagged by a cheapy data-centre.

As in, it is not CCP's fault per se.

I find coders are weak on hardware usually. I'm a hardware man; I'm weak on code.

Those inter-continental data pipes are built like tanks, with multiple fail-safes too ( as the money involved at that scale is huge ). It won't be one of them.

I R engineer *shrugs*.
Fer'isam K'ahn
SAS Veterinarians
#35 - 2016-10-16 15:26:47 UTC
Still getting socket closed ever hour, no change with DT.
Gizzie Haslack
4249003
#36 - 2016-10-16 18:31:41 UTC
Gizzie Haslack wrote:
Bedja wrote:
Kattara wrote:
Ok, so it seems that the new launcher has some nice diagnostics built-in. So the issue appears to be here:

---
K:\Program Files (x86)\CCP\EVE\Shared\tq\code.ccp\requests\packages\urllib3\util\ssl_.pyj:318: SNIMissingWarning: An HTTPS request has been made, but the SNI (Subject Name Indication) extension to TLS is not available on this platform. This may cause the server to present an incorrect TLS certificate, which can cause validation failures. You can upgrade to a newer version of Python to solve this. For more information, see https://urllib3.readthedocs.org/en/latest/security.html#snimissingwarning.
.
.
.
K:\Program Files (x86)\CCP\EVE\Shared\tq\code.ccp\requests\packages\urllib3\util\ssl_.pyj:122: InsecurePlatformWarning: A true SSLContext object is not available. This prevents urllib3 from configuring SSL appropriately and may cause certain SSL connections to fail. You can upgrade to a newer version of Python to solve this. For more information, see https://urllib3.readthedocs.org/en/latest/security.html#insecureplatformwarning.
---



Fixed the problem by deleting everything in shared cache folder still getting errors in log. But client is stable



I cleared the cache and still get dumped off of the server after 1 hour of gameplay.

Logging in is fine. It's just staying logged on for more than an hour where I have the problem :/








Quick update.


Much as the game flies like a brick a cheapy VPN is allowing me to stay connected for over an hour. A cheap SLOW VPN.

You can't fly fleet etc when the game is like a brick. Slow connections are not an option.



So fast connections like my Fibre drop, but slow ones are ok. That's the datacentre. One of their switches is going. Slow connections have more error tolerance than fast ones.

Either that or they are deliberately starving you ( CCP ) of bandwidth or processor cycles, which I hope isn't the case.
Fer'isam K'ahn
SAS Veterinarians
#37 - 2016-10-16 19:09:51 UTC  |  Edited by: Fer'isam K'ahn
So, I tried a few more things, remembering from other games that local cash can sometimes be responsible.

I deleted all cash files.

The single instance PC did fine (did °°)
The multi instance PC still lost socket on the first instance (so I adjusted fleet) every 60 minutes sharp.

Now at 18:56 (which is a full 60 minutes rotation since I started the launchers) my single instance PC also lost socket again, after 5 hours, but sharp on the 60 minutes mark.

My multi instance PC lost socket now on a random toon... looks like I am in a 'lost socket shuffle' there now (has been the same toon, always the first, for 5 hours). Two instances (toons) running on the same machine are still unaffected, which is very strange (since its the same install on the same machine with the same connection.)

Edit: LOL, now, for the first time, just while writing, a second toon dropped out of the 60 minutes rotation ??? And I dropped the same time earlier as Pandemic Legion 'allegedly' did ... but that was probably just a coincidence ,)
Alebrelle Kuatu
Sebiestor Tribe
Minmatar Republic
#38 - 2016-10-17 03:14:07 UTC
Looks to me as though it probably isn't CCP's fault on this one.

https://www.reddit.com/r/Eve/comments/57v9jq/socket_closed_i_think_i_figured_it_out/
Gizzie Haslack
4249003
#39 - 2016-10-17 06:19:28 UTC
Alebrelle Kuatu wrote:
Looks to me as though it probably isn't CCP's fault on this one.

https://www.reddit.com/r/Eve/comments/57v9jq/socket_closed_i_think_i_figured_it_out/





Yes, but CCP have to raise the alert that there is a fault somewhere in the 'cloud'. So it is them.


I am working on a gamer-end fix using this software:

https://netbalancer.com


Like how you tune a home Broadband connection ( on a copper cable ) if you are getting a lot of errors then slow it down a bit.

Limiting exe.file to 350 kb/s up & down gives me a stable server connection, but clunky gameplay. I'm trying 700 kb/s at present. 1 MB/s was too fast for me, and I was dropped after 1 hour. 4mb/s just would not fly.


The game client tries to connect at the maximum speed. That is not always a good thing. Especially in a TCP-heavy game like EVE.

Dungeons & Dragons Online has a throttle built into the game client that allows you to control the speed that the client connects at, so CCP could build a fix here and fit it into the Launcher ( for example ).
Razgar Patnovich
Iron Whales
Goonswarm Federation
#40 - 2016-10-17 12:47:00 UTC
I was finally able to login this morning. FYI: Couldn't even open a ticket because of the networking issues.
Previous page123Next page