|I'm not sure if Cisco uses the same terminology as the switches we're using here at my work, but in this case, on the switch side, I would create a LAG for the four server connections on our switch(es).|
LAG = Link Aggregation Group
The definition of the acronym is pretty self explanatory. This allows you to aggregate your 4 connections so they behave as one on the switch side.
I will point out that, if you have a single switch, you still have a single point of failure. With all 4 members of the LAG plugged into a single switch, if the switch fails, so does communication to the server. For true redundancy, you should have multiple switches and have the LAG spread across them.
My "Server" stack is a 5 switch stack connected across the backplane using cascade cables. This allows all 5 switches to behave as a single unit. In this case, I would put one member of the LAG on a separate switch....say unit's 1, 2, 3, 4. This way, should any one switch fail, the other 3 continue running. (It's worth noting I have dual uplink's in the stack on units 1 and 4 so should any switch in the stack fail, the rest of the stack is still able to communicate with the rest of the network)
Alternatively, if you have two switches, even if they're not stacked, you could create a 2 port LAG on each switch and connect two of the links to each.
It matters not how straight the gate,
How charged with punishments the scroll,
I am the master of my fate;
I am the captain of my soul.