From: Mike Taylor (mike.taylor@xxxxxxxx)
Date: Thu Jun 06 2002 - 15:38:36 GMT-3
Hi everyone,
Here's an example of a scenario I encounter quite frequently. It is in
regards to my interpretation of the requirement.
While working on a practice lab, I was asked "On R7, inject into BGP all
routes learned via OSPF type-5 LSAs.....[and] never advertise these routes to
R5." Topology is as follows:
R8 ---ebgp--- R7 ---ebgp--- R5 ---ibgp ---> "rest of the lab routers."
My solution: tag the routes with a community-ID during injection from OSPF to
BGP and filter these tagged routes via route-map to peer R5. This worked just
fine. You should also know that in the lab, there were never any routes
injected into BGP from R8, and the only routes injected into BGP from R7 were
these OSPF routes.
After lab completion, I checked to see how I did on everything and was
surprised to find their solution to the same problem: access-list 1 permit
any, route-map filter deny 10 - match ip address 1, neighbor R5 route-map
filter out. They denied all possibility of any routes passing via BGP from R7
to R5.
When I perform these labs, I find myself providing solutions that will allow
the network to "grow" - even if there might be no need for it in a particular
lab. Concerning the actual test (which for me is coming up in a few weeks),
would you recommend I take the "satisfy the lab requirements. period."
approach as they did in my scenario? Or, is it acceptable to provide a
solution such as I did? Opinions welcome.
Mike
This archive was generated by hypermail 2.1.4 : Tue Jul 02 2002 - 08:12:27 GMT-3