RE: OSPF P2M non-broadcast interface issues

From: Michael Stout (michaelgstout@hotmail.com)
Date: Sat Jul 15 2006 - 17:49:17 ART


Just one more thought related to the broadcast keyword.

If you cant get this network to pass routes, you may want to consider
using virtual-template interfaces.

  --------------------------------------------------------------------

  From: "Michael Stout" <michaelgstout@hotmail.com>
  Reply-To: "Michael Stout" <michaelgstout@hotmail.com>
  To: deckland@gmail.com, ccielab@groupstudy.com
  Subject: RE: OSPF P2M non-broadcast interface issues
  Date: Sat, 15 Jul 2006 13:44:24 -0700
  I think your ospf will improve if you add the following comamnd to
  the
  serial interface

  ip ospf priority 255

  If you are trying to peer R4 to R5 you will need the ebgp multihop
  command on your neighbor statements.

  --------------------------------------------------------------------

  From: "Radoslav Vasilev" <deckland@gmail.com>
  Reply-To: "Radoslav Vasilev" <deckland@gmail.com>
  To: "Cisco certification" <ccielab@groupstudy.com>
  Subject: OSPF P2M non-broadcast interface issues
  Date: Sat, 15 Jul 2006 14:36:28 +0100
  Hi Group,

  This is mostly for the IEWB Team to check, but any other opinion
  appreciated.

  Currnetly working on IEWB Lab 16, task 5.1 which is BGP
  configurations.
  One of my adjacency woldn't come up (eBGP session over FR cloud), due
  to TTL
  expiration (no ebgp-multihop configured as the neighbors are directly
  connected).
  Starting from this issue, i went back to my OSPF configuration. First
  the
  setup:

  R3, R4 and R5 in a hub-and-spoke toplogy with R3 being the hub. All
  devices
  use physical interfaces and the requirenments say no broadcast for
  the
  frame-relay maps(no inverse arp as well). Common IP subnet over this
  toplogy. My configuration:

  R3, the hub:
  interface Serial1/0
  ip address 154.1.0.3 255.255.255.0
  encapsulation frame-relay
  ip ospf network point-to-multipoint non-broadcast
  clock rate 64000
  frame-relay map ip 154.1.0.4 304
  frame-relay map ip 154.1.0.5 305
  no frame-relay inverse-arp

  R4, a spoke:

  interface Serial0/0
  ip address 154.1.0.4 255.255.255.0
  encapsulation frame-relay
  ip ospf network point-to-multipoint non-broadcast
  clock rate 64000
  frame-relay map ip 154.1.0.3 403
  frame-relay map ip 154.1.0.5 403
  no frame-relay inverse-arp

  R5, a spoke:
  interface Serial0/0
  ip address 154.1.0.5 255.255.255.0
  encapsulation frame-relay
  ip ospf network point-to-multipoint non-broadcast
  clock rate 64000
  frame-relay map ip 154.1.0.3 503
  frame-relay map ip 154.1.0.4 503
  no frame-relay inverse-arp

  I have connectivity between the devices and now i need to configure
  OSPF
  adjacencie. Pasting only the configuration of R3:

  router ospf 100
  router-id 150.1.3.3
  log-adjacency-changes
  area 3457 filter-list prefix area3457in in
  timers throttle spf 4000 10000 90000
  network 150.1.3.3 0.0.0.0 area 3457
  network 154.1.0.3 0.0.0.0 area 3457
  network 154.1.38.3 0.0.0.0 area 38
  network 0.0.0.0 255.255.255.255 area 0
  neighbor 154.1.0.4 cost 97
  neighbor 154.1.0.5 cost 195

  As shown above, R3 is manually configured with R4 and R5 as OSPF
  neighbors
  over what is configured as P2M non-broadcast network (because of
  another lab
  requrement the cost to the two neighbors are not equal).

  My troubles begin when i configure the two neighbors on R3 with
  different
  costs, as then the FR interface address of R5 (154.1.0.5) starts to
  be seen
  through R4 (from R3 perspective). This is because R4 and R5 have
  another
  OSPF link, over which R4 learns about R5's FR interface. Because now
  R3
  routes through R4 for the R5's address:

  Rack1R3#sh ip route 154.1.0.5
  Routing entry for 154.1.0.5/32
  Known via "ospf 100", distance 110, metric 98, type intra area
  Last update from 154.1.0.4 on Serial1/0, 00:07:06 ago
  Routing Descriptor Blocks:
  * 154.1.0.4, from 150.1.5.5, 00:07:06 ago, via Serial1/0
  Route metric is 98, traffic share count is 1

  the eBGP won't come up.

  At the same time I don't see anything in the solultion that would
  overcome
  this.

  Rado

  _______________________________________________________________________
  Subscription information may be found at:
  http://www.groupstudy.com/list/CCIELab.html

  _______________________________________________________________________
  Subscription information may be found at:
  http://www.groupstudy.com/list/CCIELab.html



This archive was generated by hypermail 2.1.4 : Tue Aug 01 2006 - 07:13:47 ART