If it's critical for production, then you should assign some assets and efforts to it. That includes spending time examining and documenting the network architecture and the diagnostic capabilities of every element.
It should be self-evident that there is not a "do not malfunction after a decade in service" checkbox in the ControlLogix or PanelView Plus software.
The questions I recommend starting with are:
Describe the exact symptoms of a "data freeze" event.
Do all ten of the PV+ units undergo this failure at the same time ?
Are there network elements like switches or uplink cables that could be responsible for the failures ?
What error messages do the PV+ terminals show in their FactoryTalk Diagnostic histories ? It may be useful to set all the PV+ terminal clocks to the correct local time so that such histories can be compared.
How long is communications interrupted, or the "data frozen" ?
Do any other network communications interruptions on the networks that include the PV+ terminals occur, such as I/O faults or inter-controller data exchanges ?
The ControlLogix ethernet modules have some very useful information about uptime, connections, and CPU utilization built into their diagnostic webpages. It may be helpful to examine or record this information periodically, especially in the immediate aftermath of a "data freeze" event.