eth0 randomly goes down

Any ideas on how to determine why an Ethernet interface goes down randomly?

Hardware is a Dell R320 from late 2015 running Debian 8.2 on an isolated subnet running a media application. syslog shows "link is disconnected" whether I ifdown eth0 or shutdown the switchport serving it, but the latter also produces a log entry "NIC Copper Link is Down" which I don't see when this failure occurs.

There is no crontab, anacrontab, nor systemd timer entry that appears to have anything to do with the network interface.

It did it once in early December, and now 4 or 5 times in the past week. Am I watching a hardware network interface on a brand new machine go south?

I'm out of ideas to investigate and this is a really weird problem to troubleshoot. The server has 2 network ports on it so my next brilliant move is to ask the customer to move the cable to the other unused port while I reconfigure the network via the iDRAC interface, as painful as that can be.

Any pointers or ideas are appreciated.

Thanks.