FastX continues to be unreliable for most users. Even after a cold reboot of the entire cluster last Friday, most connections are getting rejected, seemingly nondeterministically and accompanied by any one of a few recurring error messages.
We still don't know the underlying cause and have contacted StarNet to see if they have any ideas. We haven't made any changes to the FastX since last semester, nor has the number of users on each host increased significantly.
In the meantime, people looking to access our systems remotely are encouraged to try using plain ssh, with or without X forwarding. We will update this post as we learn more.