Re: Dual OSPF Issue

From: Varghese Thomas (vnthomas3@hotmail.com)
Date: Wed Dec 03 2003 - 19:10:29 GMT-3


Hello Brian,

Oops......I am so sorry to miss the critical portion of the email.

It works fine with admin distance method you provided; Thanks so much, Brian.

R1#sip | in 0.0.0.0
Gateway of last resort is 10.2.2.2 to network 0.0.0.0
     10.0.0.0/24 is subnetted, 2 subnets
O*E2 0.0.0.0/0 [115/200] via 10.2.2.2, 00:01:21, FastEthernet0/0

R1#sip | in 0.0.0.0
Gateway of last resort is 192.168.1.2 to network 0.0.0.0
     10.0.0.0/24 is subnetted, 2 subnets
O*E1 0.0.0.0/0 [110/11] via 192.168.1.2, 00:00:02, FastEthernet0/1
R1#

Tx n Rd
  ----- Original Message -----
  From: Brian Dennis
  To: 'Varghese Thomas' ; ccielab@groupstudy.com
  Sent: Wednesday, December 03, 2003 2:14 PM
  Subject: RE: Dual OSPF Issue

  Did you see this part of my e-mail?

  <Quote>
  Since the two OSPF processes do not communicate with each other, the only
  way to prefer a route from one process over another is to change the
  administrative distance of the OSPF process itself.
  </Quote>

  You are trying to prefer a route based on the cost. This will not
  have the desired effect. Change the administrative distance of the OSPF
  process itself. In my example I changed the administrative distance of
OSPF
  to prefer one route over another as the router does not compare the cost
  between processes.

  Brian Dennis, CCIE #2210 (R&S/ISP-Dial/Security)
  bdennis@internetworkexpert.com
  Toll Free: 877-224-8987
  Direct: 775-745-6404 (Outside the US and Canada)
  Internetwork Expert, Inc.
  http://www.InternetworkExpert.com

  ________________________________________
  From: Varghese Thomas [mailto:vnthomas3@hotmail.com]
  Sent: Wednesday, December 03, 2003 11:10 AM
  To: Brian Dennis; ccielab@groupstudy.com
  Subject: Re: Dual OSPF Issue

  Hello Brian,

  Thanks a lot for the valuable info; even after adding metric, it still does
  not prefer 'preferred route.

  BB1 - 192.168.1.0/24 R1 - 10.2.2.0/24 R2 - 172.16.1.0/24 BB2

  Both BBs sending default-route E1 type which I have no control.

  R1#sh run | be router
  router ospf 64
  log-adjacency-changes
  network 192.168.1.0 0.0.0.255 area 0
  !
  router ospf 1
  log-adjacency-changes
  network 10.2.2.0 0.0.0.255 area 0
  default-information originate metric 100

  R2#sh run | be router
  router ospf 64
  log-adjacency-changes
  network 172.16.1.0 0.0.0.255 area 0
  !
  router ospf 1
  log-adjacency-changes
  network 10.2.2.0 0.0.0.255 area 0
  default-information originate metric 200

  Event 1 - when both hubs sending default-route: R1 prefers default-route
  learned by BB1
  R1#sh ip os neighbor

  Neighbor ID Pri State Dead Time Address Interface
  172.16.1.1 1 FULL/DR 00:00:32 10.2.2.2
  FastEthernet0/0
  192.168.1.2 1 FULL/DR 00:00:31 192.168.1.2
  FastEthernet0/1
  R1#

  R1#sip | in 0.0.0.0
  Gateway of last resort is 192.168.1.2 to network 0.0.0.0
  10.0.0.0/24 is subnetted, 2 subnets
  O*E1 0.0.0.0/0 [110/11] via 192.168.1.2, 00:04:24, FastEthernet0/1

  Event 2 - when BB1 stops sending default-route: R1 prefers default-route
  learned by BB2 via R2

  R1#sion

  Neighbor ID Pri State Dead Time Address Interface
  172.16.1.1 1 FULL/DR 00:00:35 10.2.2.2
  FastEthernet0/0
  R1#sip | in 0.0.0.0
  Gateway of last resort is 10.2.2.2 to network 0.0.0.0
  10.0.0.0/24 is subnetted, 2 subnets
  O*E2 0.0.0.0/0 [110/200] via 10.2.2.2, 00:00:12, FastEthernet0/0
  R1#

  Event 3 - when BB1 re-sending default-route: R1 still prefers default-route
  learned by BB2 via R2, even though it has higher metric and E2 type.

  R1#sip | in 0.0.0.0
  Gateway of last resort is 10.2.2.2 to network 0.0.0.0
  10.0.0.0/24 is subnetted, 2 subnets
  O*E2 0.0.0.0/0 [110/200] via 10.2.2.2, 00:02:42, FastEthernet0/0
  R1#sh ip os da ex
  R1#sh ip os da external

  OSPF Router with ID (10.1.1.1) (Process ID 1)

  Type-5 AS External Link States

  Routing Bit Set on this LSA
  LS age: 794
  Options: (No TOS-capability, DC)
  LS Type: AS External Link
  Link State ID: 0.0.0.0 (External Network Number )
  Advertising Router: 172.16.1.1
  LS Seq Number: 80000004
  Checksum: 0x63C5
  Length: 36
  Network Mask: /0
  Metric Type: 2 (Larger than any link state path)
  TOS: 0
  Metric: 200
  Forward Address: 0.0.0.0
  External Route Tag: 1

  OSPF Router with ID (192.168.1.1) (Process ID 64)

  Type-5 AS External Link States

  Routing Bit Set on this LSA
  LS age: 1025
  Options: (No TOS-capability, DC)
  LS Type: AS External Link
  Link State ID: 0.0.0.0 (External Network Number )
  Advertising Router: 192.168.1.2
  LS Seq Number: 80000008
  Checksum: 0x8A35
  Length: 36
  Network Mask: /0
  Metric Type: 1 (Comparable directly to link state metric)
  TOS: 0
  Metric: 1
  Forward Address: 0.0.0.0
  External Route Tag: 1

  R1#

  Thanks again for the support on this.

  Tx n rd

  ----- Original Message -----
  From: Brian Dennis
  To: 'Varghese Thomas' ; ccielab@groupstudy.com
  Sent: Tuesday, December 02, 2003 8:25 PM
  Subject: RE: Dual OSPF Issue

  Dual OSPF processes can be messy to say the least. When I was at
  Cisco in 1996, I supported a large customer that ran dual OSPF processes in
  a few locations. Needless to say their NOC had a hard time troubleshooting
  OSPF problems in these locations and ended up calling me a lot ;-)

  Since the two OSPF processes do not communicate with each other, the
  only way to prefer a route from one process over another is to change the
  administrative distance of the OSPF process itself. Back in the old days
  ;-), around IOS version 10.3, this was not an option ;-) Support for
  comparing the administrative distance between OSPF processes was added in
  IOS version 11.0.

  In this example, two routers (R3 & R5) are sending a default route
  as an OSPF type-E2 to R4. R3 is sending the default with a cost of 30. R5
  is sending the default with a cost of 25. Theoretically R4 should prefer
  R5's default as it has the lower cost. But as we can see R4 is choosing the
  route with the higher OSPF cost. This is happening because the two OSPF
  processes are no communicating with each other. By increasing the
  administrative distance of OSPF process 1 (default from R3), we will start
  preferring the default learned from OSPF process 2 (default from R5).

  Rack2R4#sho run | be router ospf
  router ospf 1
  router-id 150.1.4.4
  network 10.34.34.4 0.0.0.0 area 0
  !
  router ospf 2
  router-id 150.2.4.4
  network 10.45.45.4 0.0.0.0 area 0
  !
  Rack2R4#show ip route ospf
  O*E2 0.0.0.0/0 [110/30] via 10.34.34.3, 00:11:36, Ethernet0/0
  Rack2R4#conf t
  Enter configuration commands, one per line. End with CNTL/Z.
  Rack2R4(config)#router ospf 1
  Rack2R4(config-router)#distance 115
  Rack2R4(config-router)#^Z
  Rack2R4#show ip route ospf
  O*E2 0.0.0.0/0 [110/25] via 10.45.45.5, 00:00:03, Ethernet0/1
  Rack2R4#

  Brian Dennis, CCIE #2210 (R&S/ISP-Dial/Security)
  bdennis@internetworkexpertcom
  Toll Free: 877-224-8987
  Direct: 775-745-6404 (Outside the US and Canada)
  Internetwork Expert, Inc.
  http://www.InternetworkExpert.com

  -----Original Message-----
  From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
  Varghese Thomas
  Sent: Tuesday, December 02, 2003 9:13 AM
  To: ccielab@groupstudy.com
  Subject: Re: Dual OSPF Issue

  Hello,

  I forgot to add the folowing

  1. Tried ospf distance ext 115
  2. tried distance 90 192.168.1.2 0.0.0.0
  ----- Original Message -----
  From: Varghese Thomas
  To: ccielab@groupstudy.com
  Sent: Tuesday, December 02, 2003 10:58 AM
  Subject: Dual OSPF Issue

  Hello,

  Need help on a dual OSPF issue.

  Setup is as follows:

  BB1 - 192.168.1.0/24 R1 - 10.2.2.0/24 R2 - 172.16.1.0/24 BB2

  Requirement:

  1. R1 & R2 should only pass default-route learned from BB1 & BB2
  respectively
  2. If BB1 stops sending default-route, R1 should have default-route from
  BB2
  via R2; vice vesa.
  3. When all are working, both BB1 is sending external default-route, type
  1.

  BB1 & R1 is running OS process 64, R1 & R2 is running OSPF process 1; R2 &
  BB2
  is ruuning OSPF process 64.

  R1's Config:

  router ospf 1
  log-adjacency-changes
  network 10.2.2.0 0.0.0.255 area 0
  default-information originate
  !
  router ospf 64
  log-adjacency-changes
  network 192.168.1.0 0.0.0.255 area 0

  R2's config:
  router ospf 1
  log-adjacency-changes
  network 10.2.2.0 0.0.0.255 area 0
  default-information originate
  !
  router ospf 64
  log-adjacency-changes
  network 172.16.1.0 0.0.0.255 area 0

  R1's RT

  R1#sip | in 0.0.0.0
  Gateway of last resort is 10.2.2.2 to network 0.0.0.0
  10.0.0.0/24 is subnetted, 1 subnets
  O*E1 0.0.0.0/0 [110/11] via 192.168.1.2, 00:00:12, FastEthernet0/1

  R2's RT:

  R2#sh ip route | in 0.0.0.0
  Gateway of last resort is 172.16.1.2 to network 0.0.0.0
  10.0.0.0/24 is subnetted, 1 subnets
  O*E1 0.0.0.0/0 [110/11] via 172.16.1.2, 00:10:49, FastEthernet0/0

  However, if I bring down, r1-bb1 link or r2-bb2 link, and brings back,
  things
  are different.

  R1' RT:

  Gateway of last resort is 10.2.2.2 to network 0.0.0.0
  10.0.0.0/24 is subnetted, 2 subnets
  O*E2 0.0.0.0/0 [110/1] via 10.2.2.2, 00:00:26, FastEthernet0/0

  R1 prefers E2 route than a E1 route:

  If I make route-metric type to E1, it still prefers higher metric route:

  R1#sip | in 0.0.0.0

  Gateway of last resort is 10.2.2.2 to network 0.0.0.0
  10.0.0.0/24 is subnetted, 2 subnets
  O*E1 0.0.0.0/0 [110/20001] via 10.2.2.2, 00:14:54, FastEthernet0/0

  R1' OSPF DB:

  R1#sh ip os da ex 0.0.0.0

  OSPF Router with ID (192.168.1.1) (Process ID 64)

  Type-5 AS External Link States

  LS age: 2703
  Options: (No TOS-capability, DC)
  LS Type: AS External Link
  Link State ID: 0.0.0.0 (External Network Number )
  Advertising Router: 12.12.12.51
  LS Seq Number: 80000002
  Checksum: 0x6179
  Length: 36
  Network Mask: /0
  Metric Type: 1 (Comparable directly to link state metric)
  TOS: 0
  Metric: 1
  Forward Address: 0.0.0.0
  External Route Tag: 1

  Routing Bit Set on this LSA
  LS age: 1216
  Options: (No TOS-capability, DC)
  LS Type: AS External Link
  Link State ID: 0.0.0.0 (External Network Number )
  Advertising Router: 192.168.1.2
  LS Seq Number: 80000001
  Checksum: 0x982E
  Length: 36
  Network Mask: /0
  Metric Type: 1 (Comparable directly to link state metric)
  TOS: 0
  Metric: 1
  Forward Address: 0.0.0.0
  External Route Tag: 1

  OSPF Router with ID (10.1.1.1) (Process ID 1)

  Type-5 AS External Link States

  Routing Bit Set on this LSA
  LS age: 978
  Options: (No TOS-capability, DC)
  LS Type: AS External Link
  Link State ID: 0.0.0.0 (External Network Number )
  Advertising Router: 172.16.1.1
  LS Seq Number: 80000001
  Checksum: 0x185C
  Length: 36
  Network Mask: /0
  Metric Type: 1 (Comparable directly to link state metric)
  TOS: 0
  Metric: 1
  Forward Address: 0.0.0.0
  External Route Tag: 1

  Thanks in advance.

  Tx n RD

  _______________________________________________________________________
  Please help support GroupStudy by purchasing your study materials from:
  http://shop.groupstudy.com

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

  _______________________________________________________________________
  Please help support GroupStudy by purchasing your study materials from:
  http://shop.groupstudy.com

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



This archive was generated by hypermail 2.1.4 : Sat Jan 03 2004 - 08:25:35 GMT-3