From: Johnny Dedon (johnny.dedon@xxxxxxxxxx)
Date: Thu Feb 15 2001 - 20:25:36 GMT-3
Damien,
While I am not yet a CCIE, I say that the path that the ip packet takes will
be as direct as the spf process has calculated, ie it will not be forwarded
through the transit to the next hop and then back to the destination. The
transit area is required for the lsa process to work. The routers, after
receiving all their appropriate lsas, and running spfs will forward routed
packets by direct paths.
Johnny Dedon
Senior Staff Consultant
Exodus Professional Services
johnny.dedon@exodus.net
www.exodus.net
----- Original Message -----
From: "Damien" <damien@clara.co.uk>
To: <ccielab@groupstudy.com>
Sent: Thursday, February 15, 2001 10:08 AM
Subject: OSPF Guru's
> Just thought I would drop this in as I have not had time to try it in =
> the Lab........
>
> If you have your typical Virtual Link set up....for a disconnected Area =
> i.e.......using another non Backbone Area as a Transit to the Area =
> 0..........
>
> | |
> | |
> Area 2 | Area 1 | Area 0
> | |
>
> so you have a virtual link between Area 2 & 1's ABR and Area 1&0's ABR.
>
> If traffic from Area 2 is destined to Area 1 what path will the packet =
> take. From what I understand in RFC 2328 Moy, the Virtual link enables =
> Transit Traffic to be forwarded through Area 1, but the actual path the =
> transit data traffic takes does not follow the virtual link. In other =
> words, virtual links allow transit traffic to be forwarded through an =
> area, but do not dictate the precise path that the traffic will take.
>
> If you read Section 16.3 of the RFC, it seems to indicate that the =
> Traffic, based on the way the SPF is calculated will go straight from =
> Area 2 into Area 1. The common misconception being you have to Traverse =
> through Area 1 and then to the Virtual link next hop i.e. Area 0 before =
> you get to the Inter-Area route, when it appears this is not the =
> case.....
>
> Can anyone who has confirmed this let me know, or knows of any other =
> Document that details this other than the RFC .....from all the people =
> (CCIE's) I have have asked some have confirmed......although with =
> reservation that they think this is the case.....?
>
> Damien
> CCIE#6634
>
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 10:28:50 GMT-3