Normally always ISP / Route related. Rarely CCPs side.

True. However, it has been CCP's side (or very close) lately - because many times when I log back on with my Jita trade alt, there's almost nobody in local.

Traceroute also shows the problem is the last hop before ccp's servers.

Put a stick in it so it cant close.

*coughcoughcough*

I believe the correct vernacular is:

:giggity:

It's not CCP, here is a brief explanation: https://forums.eveonline.com/default.aspx?g=posts&m=5022181#post5022181

You know that? What even suggests that? Besides the sheep mentality on this forum in regards to this issue. Please do present the evidence which has allowed you to conclude this.

Here is what we know:

- There are many players having the exact same issue.
- The issue started coming up for the majority of these players around the same time.
- The players reporting the issue are in different parts of the world.
- No one seems to be reporting issues staying connected to any other server.
- When logging back in, large chunks of players have also lost connection.

Whether it is CCP's servers or not, it is certainly on CCP to figure this out. To look at the evidence at hand and conclude it's on the users or their ISPs to do so is idiotic.

Have you tried the Grrr goons thread ?

I seem to recall that the message used to give more information, IIRC it read something like this:
Your account failed a heartbeat check and will be restarted

Which made it a little less likely that hundreds of threads would appear on the subject.

(Assuming the same sort of error)

I've had zero issues past weeks connecting to the CCP servers. Western Europe (Portugal) here btw

With that fact, the "many players" probably were routed thru a path that has been having issues for a few weeks back. Also considering the time gap this happens (from forum posts) It might be an issue with hops that get congested on the 11pm/1am UTC bracket?

Honestly something similar happened to me with Steam, where I could login but for some reason couldn't access the store. It happened to me and probably very few others since I searched the interwebs for this issue and there wasn't much to work around. Several visual tracerts with pingplotter later, I pinned down to one of their CDN servers from akamai (not steam itself tho). Later on, issues were solved. I now connect fine to steamstore, since Claranet, the hub before the content server somehow "shortened" the pathway from my pc to the content server that serves my region, going from more than 10 hops to just 4.

So yes, it might be an issue not on the EVE servers, but maybe CCP *can* help out by slapping the wrist on their providers.

Also, "closed socket" errors usually happen because there's too much latency. EVE might not be that much dependant on bandwith, but it is quite sensible to network latency. One of the hops starts to bog down the connections for some reason (traffic shaping *might* be one of the reasons btw), latency starts to go up.

I wasn't affected by this at all, but is there a possibility of some free stuff getting handed out?

You know that?

It's been happening a lot more to me recently to. It's always happened a little bit, despite never happening on any other online game. But the other day for instance I was disconnected 10 times while mining within 1 hour... Two of which occurred while trying to log back into the game from another Socket was Closed.

But I guess this is another case of it being the players fault and not CCP's right?

'Some of my clients close, but not all':

Socket closed messages happen when (for whatever reason) the client does not recieve a message it's expecting, in a timely fashion.

If, for example, you're sitting in a station, doing nothing at all, you have a better chance to not see the message, if there's a brief interruption between you and CCP. Your client just wasn't expecting anything at the time. If you tried opening the market, and, say, a link was saturated somewhere between CCP and you, then you're more likely to see it.

Make sense? Eve's very light on bandwidth, sending only what is needed. It means it's a little more likely to survive very short net issues.

2) set target to TQ server. I forgot which one is it. can someone tell me the IP for it?