Since yesterday evening, I have been having server connection dropouts really regularly during a session. Just had 2 within 4 minutes of one another. My internet connection seems stable here, people can still hear me on comms.
Any ideas as to why this is happening?
Did you do a route trace to see if the entire path from your ISP to the server is stable?
Just had two within about 30 minutes this morning. No apparent issues with the rest of my Internet connection. Watching an EVE twitch streamer on my second monitor and my view of his stream nor his game connection were ever interrupted while I got the Connection Lost message on my client. Weird. Going to keep monitoring it.
Just got dropped again after I was logged back in for 5 minutes.
I too am having the same issue. No problem with connectivity for past YEAR, and suddenly the game connection is so fragile as to make the game nearly unplayable. Despite all the EVE safeties for disconnects, it’s still way too risky and/or unfair to enter any mission involving timers. I’m now at the point where I log in and hope to complete two daily missions before the ubiquitous disconnect, and retire for the night. This is totally unsatisfactory. If this keeps up, I will be asking for my subscription fees to be refunded.
No issues with any other games/Internet connectivity. This is strictly an EVE issue.
It could still be an internet issue even if only EVE is affected for you, for example if only the route you use to reach the EVE server is impacted but not the route to other game servers or twitch streams.
I see. And just exactly how do I diagnose this possible Internet issue, and more importantly, how do I “fix” it?
So tired of the game developers pointing fingers at ISPs/routing and ISPs pointing fingers at game developers.
I suspect there is an issue with EVE buffering, such that a momentary (a few microseconds) spike in ping time completely severs the game connection.
Every time I’ve encountered similar problems with other games, it was ALWAYS the game that was at fault.
BTW, attempting to play today, I was disconnected 4 times within 10 minutes, requiring nearly 3 minutes for EVE to reset to the point where it would allow me to log back in again. Totally unplayable. I can’t even complete a couple of dailies.
Last point: ANY well-produced consumer software should be able to perform well regardless of whatever network routing is available (barring drastic Internet issues/shut down), whatever computer system is used by the consumer (provided it satisfies stated hardware/software requirements), and whatever common security measures are in place (firewall, VPN, anti-virus, etc.)
The other applications/games I use do not suffer from the connection fragility I am faced with by EVE Online. I should not have to jump through hoops and diagnose possible (though unlikely) network issues that, apparently, no other MMORPG game is challenged by; many of which have far more simultaneous users than EVE Online:
World of Warcraft - no issues
War Thunder - no issues
World of Tanks - no issues
Eve Online - virtually unplayable
Let’s stop making excuses for sketchy code, and start demanding that developers FIX problems instead of pointing fingers elsewhere.
Games have different server locations. As such, your connections to the servers will take different routes.
If your ISP has trouble with the connection to one country but not to another it can just happen that one of your games is affected while the others aren’t. This doesn’t have to be caused by that game.
I’m not well known with networks aside from the basics, but I do know that your ISP is a more likely cause for the connection troubles than CCP. Especially when not everyone else who plays the same game is facing connection issues.
This issue has been affecting most everyone I have contact with in the game. We are all from different parts of the world, but at any given moment multiple clients can lose connection. If this is an issue with an ISP, it would be the ISP on CCP server’s end if not an issue with the server’s themselves.
Same here
And yet…CCP seems to have absolutely ZERO interest in fixing the constant crashing. They’re more interested in developing “new” content in hopes of attracting MORE players to the game (i.e., subscribers’ money), which will probably result in even MORE instability/crashes.
BTW, I have run TRACERT to the Tranquility server, and there are NO ISSUES with connectivity through the Internet/Interweb. What I once thought were disconnects are more likely EVE CRASHING. This is most noticeable when jumping from system to system or particularly likely when (God forbid) attempting to do two things at once, like checking inventory while warping to a gate…almost a guaranteed crash. The EVE server apparently just can’t handle it!
What do we have to do to get CCP’s attention on this???
I’m having this issue relatively recently as well. Discord members told me to do packet tracing, use a VPN, that it was suddenly my ISP because “things change suddenly”, that bad routing somehow suddenly started happening. All of this magically happening out of the blue and somehow not affecting any other game or activity except EVE.
Is there a known IP address or IP range for the EVE servers that would be handling this so I can start my testing?
I have been having this problem as well, and it started a few weeks ago out of the blue. I have tried several fixes and searched for solutions but still having the same problems. I have submitted tickets and bug reports, but no response. This is affecting a lot of people, I hope they come up with a solution soon.
Same problem, tried multiple diagnostic tools…
-First the straight WAN cable connection to the network card so as to circumvent any unwanted Router lag or bad connection there.
-Second VPN with stable connection and low latency.
-Third Static IP with port forward for the game.
All of these things and STILL the game disconnects on a daily basis, most of the time it is a matter of minutes, other after 1-2 hours of play.
No way to properly diagnose something like this…
The answer to this question is no. We can speculate a lot of different avenues. However most users will jump to the conclusion “The game must be broken.”. Even when it is the one and only online game not working for you, it could be a regional network issue and not the server or client code. It is very annoying when it happens to be a middle man issue.
Solving the middle man issue is not easy or free. What you must do is identify who the “trouble child” is in the trace route and report it to your ISP. If they take care of it, count your blessings. Otherwise they might charge a one time maintenance fee to route the DNS traffic more direct and cut out the issues. Many businesses will pay for a more direct path and yet it will reoccur again at some point in the future.
Stating there is nothing wrong with the code, in any online game, is also incorrect. User will assume the issue is not the server or client, based on the fact they don’t have random disconnects. For example; I happen to know the client has a memory issue. I have a monitor program I wrote to check the RAM used for the client. Normally it starts at 30% of my total RAM and eats its way up to 65%. If I hit above 70%, it gives me an audio alarm. At 80% used, I have it force a kill signal before it locks up my system. However this doesn’t cause the disconnect, I am preventing my system from locking up from the lack of RAM. I am not paid to solve their issues and I am certain they must be aware of this memory leak.
No code is 100% or there would be no need for version and revision updates. I may have stated this on this forum elsewhere, back in my days of programming, we didn’t depend on third party software to make our code. Back before the days of DirectX, PhysX, and other third party libraries, we had to code the graphics, physics, and all the math. These shortcuts make coding easy, but it impairs programming to almost a brainless point and click state. Today’s programmer only needs to know their part and include the library used for networking. No one appears to question, if the library excludes certain hardware, or is incompatible with proprietary drivers.
I think these are all linked…
They probably do, but it only really affects a small subset of users as it only builds up when you travel a lot, and many people have a lot of ram these days. I notice it on the Steam Deck, and only when I go from a null sec pocket in Syndicate to Jita and back (somewhere in the region of 40 jumps). I dock, quit the client, reopen it and it’s fine again (dead giveaway for a memoy leak).
Users on more powerful systems are benefitting from more ram, plus a larger page file, and the fact that Windows does manage memory a bit better than the linux/proton combination (there are multiple layers of abstraction and things get lost along the way).
They’ll never fix it, it’s been in the game for 10 years or more. I imagine the causes are multiple and not easy to find and fix. Game optimisation is hard.
I agree, people tend to assume it is just “the way” to keep adding more RAM to the system and buy a “better” computer when they can no longer add more. This is consumerism at work doing their best to keep them buying.
Yes. The memory leak is due to travel from system to system, as you said, and when I am on my little 8 GB box, Jita can crash it. Swap and page files can only handle so much. When running on a SSD, I don’t recommend using drive space as temporary RAM space. I certainly wouldn’t like to ponder how much damage is done to the poor SSD from all those read/write commands.
I know system disconnects can be caused by many different things. If these companies weren’t so paranoid, they could have virtually no issues from disconnects. With today’s internet, when a DC occurs in a business environment, the client can quickly reconnect to the server. We made each client a temp server storing data in an encrypted file then the file gets uploaded to the server when reconnection occurs. We set a time and file size limit to ensure the client would be notified, if the outage was longer than a few minutes. In one instance, back in 1998 we set the system up to just ignore the user’s connectivity unless they were productive. In short, our network ignored all the AFK workers. That saved a ton of bandwidth.