Day three of Tranquility’s connection issues look much like days two and one—an inability to connect or random disconnections—only now we have chat problems. As of 17:00 Eve time on Friday, “connection failed” is still the norm, and if you do in fact connect after hammering the login button, chat is empty in known space.
CCP hasn’t been too forthcoming about the technical issues behind these problems, limiting their few status posts to what might sound like wishful thinking to some. The sum total of CCP’s status reports since yesterday are as follows:



With the weekend upon us, and war heating up in Tenal, let’s hope CCP resolves these issues so we can all get back to playing the game.
Axhind
Can’t they just admit that the cloud is pure and utter shit and bring back the damn chat servers in house so the thing actually works when users manage to login?
January 31, 2020 at 8:41 PMJohn Newell Axhind
Pretty sure it has to do with the ddos and whatever preventative measures have been taken to fight it.
CCP makes plenty of mistakes, but this time around lets blame the bad guys. Its frustrating for CCP too.
February 1, 2020 at 5:17 AMAxhind John Newell
DDoS just makes it more obvious that having the system separated is a bad idea because then external factors can affect the connection.
If the chat servers were in the CCPs data centre together with the rest of tranquillity they would not have chat problem as that communication would be internal.
So that when the user was able to connect to TQ everything would work. Sadly Hilmar probably heard that cloud is the new hotness and fell for it. The only good thing here is that he is hardly the only one to fall for that bullshit.
February 1, 2020 at 9:43 AMchthulan Axhind
I’d imagine that cost was a more compelling factor.
February 1, 2020 at 8:25 PMAxhind chthulan
To be honest I doubt it. They already run their own cluster, a rack more or less is hardly going to be a large difference in price.
However, we do know that Hilmar is an absolute moron who never thinks further than a range of a low energy alpha particle.
February 2, 2020 at 11:53 AMchthulan Axhind
I know what you’re saying but – maybe for reasons we don’t know – the chat servers have been taking a lot more maintenance? That’s what costs, not so much the hardware or bandwidth.
February 2, 2020 at 1:42 PMAxhind chthulan
AWS diesn’t offer maintenance of software. They offer HW or VMs. It’s CCP that runs the chat code on top of it.
February 2, 2020 at 6:24 PMBest of all is that they can hardly claim that they moved it to the cloud because of scaling benefits (one of a very few true things said about the cloud) as the main scaling issues are with the actual TQ cluster and the way code works (actually there are hard mathematical limits to how much you can parallelise the code that depends on each other).
chthulan Axhind
Ah, thanks, I didn’t know the chat servers were on AWS. What you’re saying makes sense now 🙂
February 2, 2020 at 6:31 PMJohn Newell
Known sec blackout!
February 1, 2020 at 5:18 AM