Prior to 802.1q, switch vendors used proprietary methods to manage the transfer of VLAN tagged ethernet frames between switches. These methods added the vlan to the frame creating a proprietary frame. Otherwise the other switch would put the frame in the vlan it’s port was configured for.
Each switch has the following VLANs configured
In the example below prior to the proprietary methods, you had to have a separate physical cable per VLAN. This connection method would become unscalable. Each VLAN 1, 2, 3, and 4 use a separate link to connect between the switches. With over a 10 VLANs in a single switch this solution would not scale, nor would Spanning Tree Protocol (STP) allow it.
The proprietary methods allowed for vendors to use a single link to pass frames for different VLANs and maintain the separation. But different vendor’s switches did not work with each other. In this example, Cisco Inter-Switch Link (ISL) is used to encapsulate ethernet frames and add on the VLAN information.
With advent of 802.1q, all switch vendors adopted the IEEE 802.1q standard. The IEEE 802.1q tag added additional fields to the ethernet frame for carrying vlan info and class of service (CoS).
- 4094 vlans
- 8 CoS values
Think of each frame with an 802.1q header carrying a tag/flag of the vlan the frame belongs to. In the 802.1q example below the frames are sent with information inside (represented as colored triangle flags) . The default vlan 1 is sent untagged/native and does not contain an 802.1q tag, thus no flag.
<-PREVIOUS Why do we use VLANs? NEXT-> Why use link aggregation?