cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1558
Views
5
Helpful
5
Replies

Q in Q / dot1q tunnel

rdubo
Level 1
Level 1

Hi,

I would like to be sure I understand Q-in-Q, and especially the configuration of this feature...

I understand that the ISP add a 802.1q tag for each customer. We can say that 1 customer = 1 vlan?

So when I configure the switch like this:

nterface FastEthernet1/0/3

description CustomerA=vlan103

switchport access vlan 103

switchport mode dot1q-tunnel

duplex full

speed 100

no mdix auto

is it correct to say that my custmer A = vlan 103?

I'm a little confused with my explanations..hope I'm right

Thanks

1 Accepted Solution

Accepted Solutions

swaroop.potdar
Level 7
Level 7

Yes your assumption is right.

One or multiple Vlans received through the interface Fa1/0/3 would be encapsulated in Vlan 103. This is the SP Vlan assigned to the customer. The actual inner Vlan could be different.

But from a Service Providers point of view, you will switch based on the identifier of Vlan 103 for the specific customer under interface Fa1/0/3.

HTH-Cheers,

Swaroop

View solution in original post

5 Replies 5

swaroop.potdar
Level 7
Level 7

Yes your assumption is right.

One or multiple Vlans received through the interface Fa1/0/3 would be encapsulated in Vlan 103. This is the SP Vlan assigned to the customer. The actual inner Vlan could be different.

But from a Service Providers point of view, you will switch based on the identifier of Vlan 103 for the specific customer under interface Fa1/0/3.

HTH-Cheers,

Swaroop

charlestran
Level 1
Level 1

In order to avoid confusion, I called VMAN. The concept and benefits are the same as native 802.1q in LAN environment but apply effectively for WAN and providers.

VMAN is for Provider and it is transparent to customers.

Make sure the provider switch set it system MTU size more than 1500 to accommodate native VLAN traffice.

Hi all,

I am not sure how the MP2MP EVC work in Metro Ethernet? As understanding, MP2MP EVC provide the multipoint-to-multipoint circuit through Ethernet network for different end-site of customer base on Provider VLAN ID. So how it can decide which end-site to forward the traffic while it has the a unique provider VLAN ID for all end-sites?

Thanks for your clarification.

MP2MP EVC works on the principle of grouping.

A group of CE's which want to access a Vlan X are trunked till your aggregation and encapsulated into a provider Vlan.

And than Provider Vlan is accessible to all such aggregation points where the QinQ is done.

The same principle of flooding and learning is used to learn all end CE's MAC's even though its using a Provider Vlan.

Hence since the MAC learning table is there, all the devices in the network know exactly where to forward trafic for a end CE even though there is Provider Vlan.

Although the above can be explained better though a diagram, still I believe this will give you a fair idea that all the PE's still need to learn all the end CE's MAC's in a ME network. Thats how forwarding happens.

HTH-Cheers,

Swaroop

Many thanks for your explanation,Swaroop.