BGP to EIGRP Redistribution Query....

Answered Question
Aug 12th, 2010
User Badges:

Dear All,


I have a doubt in BGP to EIGRP redistribution. My Topology is as Below



                           |--------------------------> A Router -------------------------------------------|

                           |                                                                                      |

Core Switch ---------                                                                                       |  MPLS Cloud    ------------------------------ Remote Branch

                           |                                                                                      |

                           |---------------------------> B Router -------------------------------------------|


Router A & B are both running BGP with Serice provider and resdistributing all the prefixes in EIGRP process.


I want the prefixes learned by core switch from Router B to be alwayz primary and prefixes from Router A to always be secondary.


I have planned the Below config for router A


Route eigrp <>

redistribution bgp <> metric

distance 175 0.0.0.0 0.0.0.0 20



IN the Above Config will all the routes redistributed from bgp into EIGRP process get the admin distance of 175???, and will this be propogated to the core switch???? where in core switch has rotues from Router B with admin distacne of 170 and routes from router A with admin distance of 175.


Will This Work?


Regards,

Ranjit

Correct Answer by Jon Marshall about 6 years 9 months ago

Rick


Good point. With the topology given that is defintely the easiest thing to do. Sometimes we can overcomplicate things


Jon

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (1 ratings)
Loading.
Jon Marshall Thu, 08/12/2010 - 02:13
User Badges:
  • Super Blue, 32500 points or more
  • Hall of Fame,

    Founding Member

  • Cisco Designated VIP,

    2017 LAN, WAN



Ranjit


There are much easier ways to achieve what you want -


1) if the remote network are summarisable then from router A advertise an EIGRP summary address on the link to the core switch using the interface command "ip eigrp summary-address ...."


That way the more specific routes are advertised from B and will be used unless the link goes down and then the summary route to A will be used.


2) if the networks are not summarisable then you can use an offset-list on router A to make the metrics of the routes received from A higher than those from B so again B will be preferred to A in normal circumstance.


Both these solutions would in my opinion be better than changing the AD.


Jon

gatlin007 Thu, 08/12/2010 - 08:10
User Badges:
  • Silver, 250 points or more

Ranjit,

Admin distance is local to the router; this attribute is not propagated to neighbors.  I agree with Jon that this technique should only be used as a last resort.

Increasing the delay on secondary path interfaces is the easiest way to influence EIGRP path selection if you want it to impact all traffic.


Very good read here:


http://www.cisco.com/en/US/tech/tk365/technologies_white_paper09186a0080094cb7.shtml



Chris

Richard Burts Thu, 08/12/2010 - 08:22
User Badges:
  • Super Silver, 17500 points or more
  • Hall of Fame,

    Founding Member

  • Cisco Designated VIP,

    2017 LAN, WAN

Jon and Chris have both given good answers that address the fact that changing the admin distance would not be an  effective solution and have proposed other solutions. But neither of them has suggested what I regard as the easiest and most effective solution. When you redistribute BGP into EIGRP you must set a metric for the redistributed routes. Just set the metric of redistribution higher on A and all routes will be preferred that come from B.


Summarization, and offset lists, and changing delay on secondary paths are all solutions that would work. But nothing is easier or more effective than changing the metric at the point of redistribution.


HTH


Rick

Correct Answer
Jon Marshall Thu, 08/12/2010 - 08:37
User Badges:
  • Super Blue, 32500 points or more
  • Hall of Fame,

    Founding Member

  • Cisco Designated VIP,

    2017 LAN, WAN

Rick


Good point. With the topology given that is defintely the easiest thing to do. Sometimes we can overcomplicate things


Jon

Actions

This Discussion