VMware - Bridging a network connection with a fixed IP in the 169 range

Question

I have a virtual machine running that I wish to connect to a physical piece of hardware via an Ethernet connection from the host machine.

This piece of hardware requires a specific IP address to be available for it to connect to - 169.254.103.1.

On my host machine I have configured the NIC to a fixed IP address of 169.254.103.1 and a subnet mask of 255.255.255.0. I then set up a new virtual network (VMNet 2), set the type as bridged and selected the NIC that my piece of hardware is connected to.

On my virtual machine I created a new virtual NIC, set the network connection type to 'custom' and selected VMnet2 as the specific network that I want it to be mapped to. I then start my control application on the virtual machine and attempt to connect to my piece of hardware, however it seems there is no communication between the virtual machine and the physical piece of hardware; no packets are sent or received and nothing significant shows up in a wireshark trace.

I have checked the firewall on the VM and the rules for the application and the ports it uses are allowed.

I have allowed the ports through the firewall on the host machine.

I am new to VMware and virtual machines in general, so all advice is welcome - am I setting up the virtual network correctly? Is there anything else I should check? Is VMware capable of handling connections like this (in the 169. address range)?

Answer 1

On my host machine I have configured the NIC to a fixed IP address of 169.254.103.1 and a subnet mask of 255.255.255.0.

Why would you do that? You don't want the host to answer as 169.254.103.1, you want your virtual machine to respond that way (based on your earlier description). If this is how you set it up, your VM probably grabbed a wholly different 169.254 (zeroconf or "link local" range) address automagically and is talking to somewhere that you aren't paying attention to (unless you set it up to try the same IP, in which case if its a windows OS it freaked and took itself offline when it realized the address was already in use), while your "physical piece of hardware" (if it's trying to talk on its own) is hitting the host OS with nothing there to respond.

The host OS shouldn't have an IP assigned to that port (or it should be something you don't plan on using). The bridged setting means that the host is as uninvolved as possible with the connectivity, it's acting like the port leads directly to the virtual. The OS in the virtual should be set to assign that IP to the NIC that's bridged to that port. The rest of the connectivity should be transparent.