Re: IE Workbook lab 3 Task 2.1 and 4.1

Hi Antonio! (Great name by the way!)

The issue here stems from the use of OSPF as the IGP on the PEs. The loopbacks possess a /24 mask. The PE generates the implicit NULL, but for the /24 connected prefix. The /32 bit OSPF network type loopback advertisement will be untagged on the adjacent devices.

We solve this issue in this case through the requirement of a /32 bit mask configured on the PE loopbacks.

You were correct in that the Deep Dive author was referring to the implicit NULL as the special label.

Great work and continue to enjoy your studies!

On Jun 15, 2010, at 5:00 AM, Antonio Saez wrote:

> Hi Anthony,
>
> Could you explain this deeper?
> The NULL label advertised for the /32 prefix I suppose is the PHP implicit
> null. If true, is not advertised for a non /32 prefix?
> What is the special label you refer?
>
> Regards and thank you in advance
>
>
> -----Mensaje original-----
> De: nobody_at_groupstudy.com [mailto:nobody_at_groupstudy.com] En nombre de
> Anthony Sequeira
> Enviado el: martes, 15 de junio de 2010 6:08
> Para: Julio Carrasco
> CC: ccielab_at_groupstudy.com
> Asunto: Re: IE Workbook lab 3 Task 2.1 and 4.1
>
> Hello Julio!
>
> Without the task in front of me, I am going to guess that R4 and R5 are PE
> routers in an MPLS L3 VPN scenario. To quote a "Deep Dive" from the very
> latest version of our WB II....
>
> "The use of a /32 prefix allows the local router to advertise a NULL label
> for the prefix. Any non-host route will generate a special label that the
> local router will use to perform an aggregate lookup based on the
> destination IP address in the packet. With the NULL label advertisement, the
> label lookup will be performed based on the VPNv4 label, in the case of an
> MPLS VPN scenario."
>
> Warmest Regards,
>
> Anthony J. Sequeira, CCIE #15626
> Senior CCIE Instructor
>
> Test your Tier 1 Knowledge
> Q: For RIPng, you can only disable split-horizon ...
> A: globally
> More Info: http://www.ietf.org/rfc/rfc2080.txt
>
> On Jun 14, 2010, at 10:19 AM, Julio Carrasco wrote:
>
>> Hello experts,
>> I am reviewing this lab, and I saw, that on Task 2.1 it says that is
> necessay
>> for Lo0 on R4 and R5, to be announced with a mask of 32 bits, so the LDP
>> peering on task 4.1 needs both loopbacks to be a /32 mask.
>> I do not understand exactly why, so could anyone please help me with that
> ?
>>
>> Thanks in advance and best regards !!!!
>>
>>
>> Blogs and organic groups at http://www.ccie.net
>>
>> _______________________________________________________________________
>> Subscription information may be found at:
>> http://www.groupstudy.com/list/CCIELab.html
>
>
> Blogs and organic groups at http://www.ccie.net
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
>
>
> Blogs and organic groups at http://www.ccie.net
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html

Blogs and organic groups at http://www.ccie.net
Received on Tue Jun 15 2010 - 08:57:27 ART