VRF for VPNv4/VPNv6

Hi All

 

I have a quick question about interpretation of a requirement.

If I have two sites for example, and I am asked to establish full connectivity between the sites prefixes, is it wrong to use different RTs and fully import/export them? Or is this considered filtering already?

This is asked in Full Scale Lab 1 Task 3.2 and the solution states to use the same RT for all VRFs:

This is a very straightforward "full mesh" route policy, where the customer routes are not filtered in any way and are freely imported into the VRF. For this scenario, we will import and export the same route-target on all PE devices.

From a technical perspective I do not see any difference between freely importing between different RTs and the same RT.

Any input is welcome.

 

BR,

Tim

Comments

  • Hi All

     

    I have a quick question about interpretation of a requirement.

    If I have two sites for example, and I am asked to establish full connectivity between the sites prefixes, is it wrong to use different RTs and fully import/export them? Or is this considered filtering already?

    In my personal opinion i would say no, as long as you fulfill the requirements of the task. The task states that you should provide a RT policy that allows the two sites to freely exchange routes between them. Then how you are going to achieve this it's your personal problem.

    I remember that for this task i used to create an RT for every single PE.

  • Hi,

     

    Thanks for the reply. I also used different RTs per PE.

     

    Best Regards,

    Tim

  • If you take this approach in the lab, not doing what you're told to do, you'll magically and miserably fail; in the lab you do what you're told to do, doesn't matter if it's a best practice or not, if it makes sense or not from a real-life deployment point of view.

     

Sign In or Register to comment.