What can be solved in 2 hours

[VisionHolder]

So we're down for 2 hours. What do you think they're actually doing to actually remedy the current problem?

I would suspect that they've been trying to rule out the potential problems. 9000 sounds like a nondescript error, and so were the errors when entering instances. These problems are happening consistently on the highest traffic servers, so it has something to do with high traffic. No doubt each maintenance required resetting servers and flushing errors from the system, which takes time. Upon bringing the systems back up, they probably made changes that helped with congestion to rule out simple server overload. Maintenance or no, devs are probably combing thousands of lines of code on both client and server software. Once they rule out congestion and code errors as the *root* cause, they will be able to make a case for hardware problems, at which point they go to their vendors. Worst case, they have to get new hardware or even change vendors.

They have brought the servers down 4 times now in space of 36 hours, and every time it was for the same reason, and every time they told us they had identified and corrected the issue...

Actually, they only said they thought they fixed it once... on the second maintenance if I remember correctly. All other maintenance posts have simply stated the time and reason, along with an apology. Don't make things up to justify people's anger.