Hi, all. My problem was finally solved with changes to our network wiring configuration. Here is the write-up from my IT team explaining exactly what they did:
------------
We performed a few upgrades simultaneously which may have contributed to the current stability. Though the Oracle upgrade and the kernel upgrade / patch may have been factors, it's our opinion that the network port was the primary culprit.
The way that the network was previously configured, each server gained network access through a different switch, which in turn were connected through other devices. The database server stores the databases on the storage cluster, mapped via NFS. By moving the database server network port to the same switch that the storage cluster users, we eliminated extra hops and hardware. Speeds for the database server increased and we stopped seeing timeouts occur.
Before change:
web server -> Switch1 -> Switch2 -> Switch3 -> database server -> Switch3 -> Switch4 -> Switch5 -> storage -> Switch5 -> Switch4 -> Switch3 -> database server -> Switch3 -> Switch2 -> Switch1 -> web server
After change:
web server -> Switch1 -> Switch2 -> Switch3 -> database server -> Switch3 -> storage -> Switch3 -> database server -> Switch3 -> Switch2 -> Switch1 -> web server
------------
I guess the moral of the story here is check your network hops if anyone is experiencing these kinds of ColdFusion/datdabase timeouts and have no idea where they are coming from.
P.S. Big props to Anit for helping me troublshoot every other possibility before I handed it off to the IT team with the conclusion "It's not ColdFusion!". Adobe Support was EXCELLENT.
