We've found the error behind the server timeouts, apparently we keep running out of RAM and swap memory, so it causes strange effects.
"Our upstream have just reloaded the Cisco Cluster.
As you may or may not of noticed, CF | UK are back in action!
CF | UK is now back online