Hello Everyone,

I am deploying two Nginx servers running on Ubuntu with heartbeat for HA/fail-over. Each server has two physical NIC interfaces.

In my firewall, I have 1:1 NAT mappings from the various public IP addresss to a NAT subnet that Nginx is listening on. The NAT interface (LAN) from the firewall is plugged into the same switch (same subnet, VLAN) as my two Nginx servers. I have the first physical NIC on the Nginx server plugged into the switch, the switch configured with a VLAN, all my upstream servers are also plugged into that same switch, and are a part of the same VLAN/subnet.

The second physical NIC in each server is dedicated for heartbeat and I'll likely just plug the server ports directly into each other, or perhaps create a 3rd VLAN to also detect a switch failure.

Does anyone see any problems with this configuration or have any suggestions for improvement?

Thanks,

Matt