Things are not adding up here. Changing the MTU (maximum transmission unit) will not affect a switch's ability to deal with VLAN (802.1q) tagged frames. There are a number of different VLAN implementations out there. The first type of VLAN is port based or local only to the switch. If you define a port on a switch say for VLAN 2, traffic entering that port will be classified for VLAN 2 and will go to any port on that switch designated for VLAN 2. But if you connect another switch which only supports port based VLANs, the VLAN definitions will not transfer over to the second switch.
The next type of VLAN used on switches and the one most used by managed/enterprise switches are tag based VLANs or its IEEE reference of 802.1q. The advent of 802.1q VLAN tagging allowed VLANs to propagate between switches and be used to maintain traffic separation and enforcement. If you create the same VLAN 2 on one switch, the traffic which traverses to the second switch will be sent out to ports which are assigned to VLAN 2. How this is done is by altering the Ethernet frame by inserting a 4 byte data field. This 4 byte data field defines the VLAN ID the Ethernet frame is to be assigned to. Here's a diagram of an Ethernet frame without and with the 802.1q header inserted into the frame:
Image may be NSFW.
Clik here to view.
If a switch doesn't support 802.1q VLAN tagging, it will think a 802.1q tagged Ethernet frame is a malformed frame and will drop it. When you adjust the MTU of an Ethernet device, you are saying the device can transmit or accept a frame of x size. The default is 1500 bytes. If you adjust it beyond this, you are increasing the size of the area labeled payload in the above diagram. So as you can see, adjusting the MTU size of a device for the Ethernet frame it can support has no bearing on support for VLANs.
There are two other VLAN constructs called private VLANs and Q in Q.
The next type of VLAN used on switches and the one most used by managed/enterprise switches are tag based VLANs or its IEEE reference of 802.1q. The advent of 802.1q VLAN tagging allowed VLANs to propagate between switches and be used to maintain traffic separation and enforcement. If you create the same VLAN 2 on one switch, the traffic which traverses to the second switch will be sent out to ports which are assigned to VLAN 2. How this is done is by altering the Ethernet frame by inserting a 4 byte data field. This 4 byte data field defines the VLAN ID the Ethernet frame is to be assigned to. Here's a diagram of an Ethernet frame without and with the 802.1q header inserted into the frame:
Image may be NSFW.
Clik here to view.
If a switch doesn't support 802.1q VLAN tagging, it will think a 802.1q tagged Ethernet frame is a malformed frame and will drop it. When you adjust the MTU of an Ethernet device, you are saying the device can transmit or accept a frame of x size. The default is 1500 bytes. If you adjust it beyond this, you are increasing the size of the area labeled payload in the above diagram. So as you can see, adjusting the MTU size of a device for the Ethernet frame it can support has no bearing on support for VLANs.
There are two other VLAN constructs called private VLANs and Q in Q.
