We are about to perform an emergency failover of the department FastX gateway, which should be complete by 12:00pm. This will mean that anyone who is presently connected to FastX will lose their session(s), along with any unsaved work.
Several users reported over the weekend that the FastX gateway was refusing new connections, which I confirmed this morning. Cleaning out orphaned X sessions did not resolve the problem (which had been sufficient on previous, similar occasions). I've so far been unable to figure out either the cause of the problem or a non-disruptive fix, so I'm going to fail the service over to the standby server while I continue investigating.
I've also performed an upgrade of FastX itself (2.1.50 to 2.3.83) on the standby server, in the hopes that the new version will be more robust.