Task 2.9 BGP Filtering

When I first looked at task - I thought that addition neighbor send-community statements were required on R1 for R1 to R5 peering and also on R4 for R4 to R5 peering to allow for the failure of R5's s0/0/0 frame relay circuit.  In this situation traffic reroutes via VLAN 45.

How wrong was this assumption.  As all of the iBGP peerings are using loopback, the peering between R1 and R5 remains established during the above failure - the BGP session simply routes through R4, so R5 always gets the prefix 172.21.1.0/24 with the community local-as.

Yet another case of over config for me [:$]

Comments

  • Yes, it is really easy to over configure. Not sure you would loose points on this though as it should not affect anything in a negative way.

  • Hi all,

       Pointless configuration means that you do not know technology are you did not pay attention on task requirements; i would say overconfiguration will cost you points, be it that you added some extra commands where not needed or optimized something when not asked for. Even Bruno, the RS program manager said this in Vegas at Cisco Live, that lots of students overconfigure stuff while not being asked for, which ultimately cause loosing points.

    All the best!

  • The good news is I had valid reason for doing this, my verification proved the error of ways!  If the iBGP peering were using interface addresses as opposed to loopbacks then my configuration would make sense, but then you never configure iBGP this way anyway.

  • Why does the SG not configure send-community on the neighbor 150.1.4.4?

  •  

    You mean on R4 itself or on R1 for R4?!

    In both cases this not needed as R4 is not an exit-point nor has to resend the community to R5 as an BGP peer if the frame-relay PVC between R5 and R1 went down. As welshdragon mentioned above in his first post, the peering between R1 & R5 will remain up if this PVC went down, using the route "though" R4 because the peering uses the loopback interfaces. So R4 is just "connecting" the the two routers in that case, and peering still remain between them. This is a result of BGP acting as TCP application.

     

    HTH,

    Bassam

Sign In or Register to comment.