2.5 Routing Loop Prevention - Another Solution?

I fully understand the solution in the guide, but did mine a different way and would like feedback if I would "get the points".

 

On R6, I set a tag of routes going into EIGRP from RIP.

 

Rack1R6#show route-map FROM_RIP
route-map FROM_RIP, permit, sequence 10
  Match clauses:
  Set clauses:
    tag 6120
  Policy routing matches: 0 packets, 0 bytes

router eigrp 100

 redistribute rip metric 1 1 1 1 1 route-map FROM_RIP

 

Then, on R3 and SW2, I block these coming back from OSPF-> EIGRP.

Rack1R3#show route-map
route-map NO_ORIG_IN_RIP, deny, sequence 10
  Match clauses:
    tag 6120
  Set clauses:
  Policy routing matches: 0 packets, 0 bytes
route-map NO_ORIG_IN_RIP, permit, sequence 100
  Match clauses:
  Set clauses:
  Policy routing matches: 0 packets, 0 bytes

router eigrp 100

redistribute ospf 1 metric 1 1 1 1 1 route-map NO_ORIG_IN_RIP

 

Thanks,

 

Comments

  •  I did just the same thing you did, just like INE instructors always say there might be more than one way to implement a task, i think was is important is that you carry out what the task requires using any solution except explicitly told not to use a particulat solution/method.

  • Did mine the same althougt I chose SW2 as the router to deny the tag on.

  • I did the EXACT same thing, down to the tag number [;)].

    Seems to be a valid solution to me.......

  • Made the same configuration.

    There's no point in tagging all routes from eig to ospf. the behaviour of EIGRP will prevent routes from leaking back to the same protocol again.

    Only external routes of EIGRP would cause problems here, and they are all coming from rip, so I tagged it and filtered out when redistributing from ospf to eigrp.

    Another thing which is weird to me, in SG there are other tags on redistribution point. Why? I think it's pointless, using the same tag shouldn't make any harm here.

     

  • I used the same method as rkosylu1. Tag on R6 then deny the tagged routes on R3 and SW2.

  • In general, tagging the route at the source is a good practice, for this task it exactly does what the task is asking for. the SG solution is valid also...

  • What about just increasing ospf external AD to 171 on both R3 and SW2?

    Since there are not other redistribution points (including redistribute connected) it fits for the tasks even though it won't talke into account if anyone starts redistributing connected (the task only ask to take into account new RIP routes)

  • at first I wanted to do the same config as Rob but then I figured I wanted to use the usual redistribution framework... :)

     

     

    config t

    route-map EIGRP_TO_OSPF deny 10
     match tag 110
    route-map EIGRP_TO_OSPF permit 20
     set tag 90
     
     
     
    route-map OSPF_TO_EIGRP deny 10
     match tag 90
    route-map OSPF_TO_EIGRP permit 20
     set tag 110
     
     
    router eigrp 100
     redistribute ospf 1 route-map OSPF_TO_EIGRP metric 1 1 1 1 1
     
    router ospf 1
     redistribute eigrp 100 route-map EIGRP_TO_OSPF subnets

     

     

     

    Hope it helps !

  • i did the exact thing. just the tag numbers were different. Since no one said that this solution is not valid. im a happy man.

  • +1 form me, but neither this nor the SG solution fix the suboptimal routing
    on eihter R3 or Sw2 - one of them will follow the OSPF route to reach the
    external routes to BB1 instead the shorter EIGRP route. Too bad we can't change
    the AD based on route tag....
  • What about just increasing ospf external AD to 171 on both R3 and SW2?

     

    This was my solution.  They won't be redistribured back into EIGRP with this and it fixes the suboptimal routing issue also.

    I know the question does not speciy to fix the sub-optimal routing, but i'm a big fan of keeping things as simple as possible, and this was the most simple solution imo

     

  • exactly the same - seems this is the popular solution for this therefore the most logical..?

     

  • exactly the same - seems this is the popular solution for this therefore the most logical..?

    There are many ways to do route redistribution. All of them are acceptable as long as it doesn't violate general restriction and task restriction. In the end, reachability is the king. Try to reach this state (called golden moment) as soon as possible when you're doing your lab.

Sign In or Register to comment.