Dynamic Trunking Protocol

Welcome back to a new article on Networking Concepts. Let me first answer the question which was asked in the previous article. The link between two switches are access and we know that on access link, frames don’t carry tags. We can use this loophole to satisfy our requirements. On Switch1, we can make f0/2 interface as a part of vlan 10 and the f0/2 interface on Switch2 as vlan 20. So, when host A initiates an icmp echo request, which in general will result in an arp request and will reach the Switch1 on the f0/1 interface. Now as it is a broadcast frame, the switch will flood it on all the interfaces which are part of vlan 10 and therefore would go out of f0/2 interface. Here comes the concept of access links, switches do not add any tags on the frame and we can use this to trick the other switch to our benefit. Once the frame is sent from f0/2 interface of Switch1, it will be received at f0/2 interface of Switch2 which is part of vlan 20. As we don’t have any tags associated, so Switch2 would think that it is coming from vlan 20 itself, and will forward the frame to all the interfaces which are part of vlan 20, and this in return would reach our final destination. This is how the hosts will communicate. Pretty interesting right!!

Now let’s start with today’s topic which is DTP(Dynamic Trunking Protocol).

It is a layer-2 protocol and is a cisco proprietory.

It is used to create trunk and access port dynamically.

By-default it is enabled on all cisco switches.

Uses multicast MAC address : 0100.0ccc.cccc

Sends messages periodically after every 30 seconds.

There are 2 modes of DTP:

Dynamic Desirable(DD) – Actively negotiates for trunk.

Dynamic Auto(DA) – Passively waits for trunk.

Ports which are connecting to end devices are always access ports. The following table provides the status of the link when ports are in different modes.

The switchport comprises of two modes: Administrative Mode and Operational Mode. Administrative modes can be Static Access, Static Trunk, Dynamic Desirable, Dynamic Auto. Operational mode can be Trunk or Access.

We can disable dtp by the command nonegotiate on the interface. The trunk encapsulation protocol also comprises of two modes: Administrative Mode and Operational Mode. Administrative modes can be Dot1q, ISL, Negotiate. Operational mode can be Dot1q or ISL If both the switches are operating in Negotiate mode, than the trunking would form with the encapsulation protocol as ISL because it is a cisco proprietory. You can’t assign two different protocols on the link, as it won’t negotiate.

We have already discussed about the native vlan in our previous article, and we can manipulate this native vlan feature on our trunk link. We also know that dot1q protocol adds a vlan tag on all the frames except the native vlan frame.

Consider the above topology with two switches, hostA and hostB. Switch1 has f0/1 interface in vlan 10 and Switch2’s f0/1 interface in vlan 20. F0/2 interfaces of both the switches are configured as trunk. If host A initiates an icmp echo request for host B, which in return will initiate an arp request. This request will reach the Switch1 on f0/1 interface and as it is part of vlan 10, it will forward the received frame on all the interfaces which are part of vlan 10 as well as the trunk port. Vlan 10 is configured as native vlan on Switch1 and vlan 20 for Switch2. The received frame would be sent without any tag and therefore when received on f0/2 interface on Switch2, Switch2 will feel that it is coming for vlan 20 as there is no tag associated with the frame. Switch2 will forward this frame to all the interfaces which belong to vlan 20. Interestingly, this manipulation will allow communication between two hosts which are in different vlan.

Thank you so much for reading this and I hope you enjoyed it!

Feel free to comment with questions or additional insights! Let’s keep learning together. #Networking #Switching #DTP

?