Figure 1:
For a long time, we had two logical sites in a single physical one. When we upgraded our voice VLAN for ‘panic’ buttons, we had to connect devices on site B to site A. These buttons were meant for purposes that site A was built for in the first place. Normally by standard for us, we already had the same voice VLAN on site B. This would cause an IP conflict on the VLAN since you have two subnets on a single VLAN.
Figure 2:
To solve this problem, I had to make a new VLAN on site B, this example being VLAN 21. The panic buttons would be connected to this VLAN on site B. If you’re familiar with VLAN’s you should be aware that VLAN21 on site B cannot work with VLAN20 on site A if it’s trunked.
Figure 3:
To solve this problem I made VLAN21 on site B connecting to site A on access port, also known as untagged.
I also did this the other way around by making the interface connecting site A to site B an access port, also known as untagged, in VLAN 20.
Doing it vice versa would basically mean you have two subnets on one single VLAN but doing that has practically no use case.
This would result in more broadcast traffic and it’s a bigger pain to manage and troubleshoot in general.