bvsnarayana03 Sun, 04/20/2008 - 21:39
User Badges:
  • Silver, 250 points or more

Native vlan is used for sending control information like STP, PagP,DTP, VTP etc. Bcoz cisco had default vlan of 1, these protocols were always allowed on the trunks.

Later with dot1q, cisco control protocols were still using vlan1. If a native vlan was changed on the switch, the cisco control protocols still worked on vlan 1. Thus they were tagged, while ieee control protocols like LAcp, mst will be sent untagged.

For a trunk to work, the native vlans on both sides of the trunk has to be same. frames of native vlan i.e. control infor foes untagged over trunk.

nikeshkotadiya Sun, 04/20/2008 - 22:00
User Badges:

thanks for your reply. As per my understanding, If you don't define any native VLAN in 802.1q it will by default take VLAN 1 as native then what is the significance of defining native VLAN on 802.1q?

The reason that - by default - the native vlan is untagged is for backwards compatibility with non-802.1q devices; namely switches. Exchanging STP with these non compatible devices was important enough to the working group that the untagged native vlan was put into the spec. Note that you *can* disable sending untagged frames on .1q trunks with the "vlan dot1q tag native" global config command. All of the standard L2 control protocols (including LACP) will still work. And why wouldn't they? After all, they work with ISL and it has no concept of a native vlan.

bvsnarayana03 Mon, 04/21/2008 - 21:40
User Badges:
  • Silver, 250 points or more

Hi, thats an interesting thing u mentioned. "Tagging a native VLAN". why wud tht be required in any environment.

On the other part, yes you are right that L2 control protocols still work but this time they'll go tagged. For cisco switches there was no concept of native vlan but vlan 1 was default & all cisco control protocol always travelled across vlan1 even in mixed environment.


This Discussion