Two 3850 Stacks connected by fiber uplink - One goes unresponsive occasionally

We have a remote office, which has 2 separate suites in the same building. There is a stack of two 3850's on the first floor, then a fiber uplink to another stack of two 3850's on the second floor. The second floor was built out years after the first floor - a consultant assisted with the setup (I am very much a generalist, hence why I'm here asking for help!).

First floor is running Version 03.03.02SE RELEASE SOFTWARE (fc2)

Second floor is running Version 03.07.04E RELEASE SOFTWARE (fc1)

Now to the issue. The second floor switch will become unresponsive to ping, initiation of a putty session, etc., but will continue to function perfectly normal otherwise. I can get to workstations on site as well as other resources in other VLANs. Solarwinds will report that the switch will stop responding, then start responding, at various intervals throughout the day. Logging set to debug doesn't show anything amiss.

The first floor stack has a VLAN IP on each VLAN in that site. The second floor stack only has an IP on one VLAN. (For example, 192.168.1.254 is the first floor stack on VLAN 1 and 192.168.1.253 is the second floor stack on VLAN 1).

The second floor stack's IP of 192.168.1.253 only becomes responsive again if I ping it from a workstation/server on that VLAN. I have a server in that VLAN, for example 192.168.1.5, and if I ping the second floor stack's IP on VLAN 1, the second floor stack starts responding to ping and I can putty to it for a period of time (usually about 5 minutes), before it goes unresponsive again. I have had a ticket open with Cisco for weeks and they are stumped.

I'm sure I'm missing details, so feel free to ask away and I will answer what I can.