The Adjacency routing protocol on MPLS vs Frame Relay

Unanswered Question
Apr 22nd, 2007
User Badges:

hi..please helps. assume there is the customer have 2 site(A and B) and use ospf. how routing protocols(ex:ospf) process establish neighbor relationship(adjancency) with another routing protocol at the site B? is it true the nature of metric calculated is changed because there is redistribution? and we never be able adjacency? could MPLs act the just like Frame Relay. is there guide/document how to implement routing protocol in mpls network? i really need how to do that. tx a lot. :)

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
Harold Ritter Mon, 04/23/2007 - 05:01
User Badges:
  • Cisco Employee,

Jimmy,


The answer depends on whether you are using l2vpn or l3vpn between the two sites. With l3vpn, the adjacency will be established with the service provider router (PE) and not with customer router (CE) in the other site.


With l2vpn (eompls or vpls) the adjacency would be between the two customer router2 just as it is today with your FrameRelay connection.



Hope this helps,

jemekeren Mon, 04/23/2007 - 18:28
User Badges:

Tx Hritter, mplsvpn 3 adjacency will be made between ce and pe.i worried whether how metric ospf affected in mpls scenario because there are intervening router (PE). will pe router additional cost/metric? between two site? i am creating like "ip ospf cost 10 on both ce router. of course by creating that metric in natural(Frame relay, ethernet, serial) the metric will be only 10. so what about MPLS? how to do that in common implementation even there is another router in between that naturally? tx a lot :)

mohammedmahmoud Tue, 04/24/2007 - 00:13
User Badges:
  • Green, 3000 points or more

Hi Jimmmy,


As far as i know when redistributing into OSPF, OSPF defaults the metric to 20 for redistributed IGP routes and 1 for BGP routes, on the other side when redistributing into BGP, the redistributed route metric is copied by default into the MED field. Accordingly you can't preserve the OSPF metric in L3VPN MPLS, as it will change to 1 by default on the second PE when redistributing back from BGP to OSPF to reach the second CE.


HTH, please rate if it does help,

Mohammed Mahmoud.


Harold Ritter Tue, 04/24/2007 - 10:30
User Badges:
  • Cisco Employee,

Redistribution of the OSPF routes into BGP and then back into OSPF on the other PE means that the routes are seen as inter-area or external by the CE. If you want the CE to see the OSPF routes as intra-area, you need to use a sham-link between the PEs.


Hope this helps,

Actions

This Discussion