From: Ccie candidate (beacc1e@yahoo.com)
Date: Sun Jan 05 2003 - 15:52:23 GMT-3
Hi Carlos and everyone,
Thanks for your reply. Actually there was a posting (in 12/29/2002) regarding this and some people answered to this problem as below and I wanted to verify this solutions. My comments to this solutions at the bottom.
####################################################################
At 08:04 PM 29/12/2002, kym blair wrote:
Jay and Yong,
When you are doing mutual redistribution between OSPF and RIP, there are
several ways to get the OSPF routes into RIP:
(1) Use RIP version 2
(2) Create a /30 secondary address on the R2-R3 link so R3 will learn the
/24 AND /30 routes (repeat for other masks)
(3) Create a tunnel between R2 and R3 with a /30 mask (repeat for
other masks)
(4) My favorite if you can't use RIPv2: Create another OSPF process on
R2;redistribute OSPF 1 into OSPF 2; add summary-address statements under
OSPF2; redistribute OSPF 1 and OSPF 2 into RIP.
##################################################################
My Lab
R4-----OSPF---R6-----RIP-----R2 (without using static red on R6 or static on R2)
Solution #1 It works.
Solution#2 I don't think this (using secondary) will help to solve OSPF to RIP red. In my lab putting /30 in R6 ---R2, which is the RIP domain, only causes the /24 learned routes in OSPF domain to be advertised (by R6) and accepted(by R2) as /30 routes in the RIP domain. Am I doing something wrong ?
Any /30 subnets learned by OSPF did not redistributed to the RIP domain.
Solution#3 It works.
Solution#4 It works.
What do you guys think ?
Kafkaf.
Carlos <cchorao@xtra.co.nz> wrote:
Kafkaf ,
The scenario you depict has vlsm/flsm issues. The following 3 docummets
explain it well.
http://www.cisco.com/warp/public/105/55.html
http://www.cisco.com/warp/public/105/53.html
http://www.cisco.com/warp/public/105/52.html
1) Net 148.9.1.0/30 will not be advertised by RIP on R6 out int s0/0
reason --> same class net's but different masks;
2) Using a secondary address with a /30 mask on int t0/0 will not help,
reason ---> same as 1.
3) Applying a secondary address with a /30 mask between R6 and R2 will not
work---> same major net as primary address, no updates sent through
secondary,longest mask applied to updates recieved.
4) Ospf reflects these destinations in R6 routing table. The problem you are
trying to solve is how to get longer vlsm routes from ospf on R6 into rip.
Carlos
Telecom New Zealand - Advanced Solutions Group : Network Design and Security
----- Original Message -----
From: "Ccie candidate"
To:
Sent: Friday, January 03, 2003 6:47 AM
Subject: OSPF to RIP redistribution
> Hi everyone,
>
> I have the following setup. I was trying to do secondary address
solution(on RIP routers) for OSPF to RIP redistribution to get the OSPF
routes with different subnets than RIP.
>
> R4(to)------(t0/0)R6(s0/0-FR) ------(s5/1-FR) R2
>
> R4 and R6 is running OSPF
> R6 and R2 is running RIP v1.
> Interface information:
> R6 s0/0 ---> 148.9.100.1/24
> R6 t0/0 ---> 148.9.1.2/30
> R4 t0 ---> 148.9.1.1/30
> R2 s5/1---> 148.9.100.2/24
>
> The following are the configs for each router.
>
> R6
>
> router ospf 54
> log-adjacency-changes
> summary-address 130.0.0.0 255.0.0.0
> redistribute rip metric 10 subnets
> network 148.9.1.2 0.0.0.0 area 0
>
> router rip
> redistribute ospf 54 metric 2
> network 148.9.0.0
>
> R2
>
> router rip
> network 148.9.0.0
>
> R4
>
> r4-ccie#sh run | be router ospf
>
> router ospf 54
> log-adjacency-changes
> network 148.9.1.0 0.0.0.3 area 0
> network 148.9.1.8 0.0.0.3 area 0
> network 148.9.4.0 0.0.0.255 area 0
> network 160.1.1.0 0.0.0.255 area 0
>
>
>
> R6#sh ip route ospf
>
> 160.1.0.0/32 is subnetted, 1 subnets
>
> O 160.1.1.1 [110/7] via 148.9.1.1, 00:06:14, TokenRing0/0
> 148.9.0.0/16 is variably subnetted, 7 subnets, 2 masks
> O 148.9.1.8/30 [110/7] via 148.9.1.1, 00:06:14, TokenRing0/0
> O 148.9.4.0/24 [110/7] via 148.9.1.1, 00:06:14, TokenRing0/0
> O 130.0.0.0/8 is a summary, 01:03:53, Null0
>
> R6#
>
> R2 learns rip updates from R6 matching subnet /24, others (such as
160.1.1.0) only major network address itself is included in the
advertisement.
>
> R2#sh ip route rip
>
> R 160.1.0.0/16 [120/2] via 148.9.100.1, 00:00:11, Serial5/1
> 148.9.0.0/24 is subnetted, 5 subnets
> R 148.9.4.0 [120/2] via 148.9.100.1, 00:00:11, Serial5/1
> R 148.9.6.0 [120/1] via 148.9.100.1, 00:00:11, Serial5/1
>
> R2#
>
> Now to learn other networks with different subnets (such as 148.9.1.0/30
and 148.9.1.8/30) I have added the secondary address to the R2(s5/1) and R6
(s0/0)
>
> R6 (s0/0)
>
> interface Serial0/0
> ip address 148.9.1.65 255.255.255.252 secondary
> ip address 148.9.100.1 255.255.255.0
>
> R2 (s5/1)
>
> interface Serial5/1
> description to (FR s1/6)
> ip address 148.9.1.66 255.255.255.252 secondary
> ip address 148.9.100.2 255.255.255.0
>
> This is (bolded entries) what happens.
> **148.9.1.0/30 and 148.9.1.8/30) did not show up in the rip routing table.
>
> R2#sh ip route rip
>
> R 160.1.0.0/16 [120/2] via 148.9.100.1, 00:00:15, Serial5/1
> 148.9.0.0/16 is variably subnetted, 8 subnets, 2 masks
> R 148.9.4.0/30 [120/2] via 148.9.100.1, 00:00:15, Serial5/1
> R 148.9.4.0/24 [120/2] via 148.9.100.1, 00:01:10, Serial5/1
> R 148.9.6.0/30 [120/1] via 148.9.100.1, 00:00:15, Serial5/1
> R 148.9.6.0/24 [120/1] via 148.9.100.1, 00:01:10, Serial5/1
>
> R2#
>
> Using "ip summary-address rip 148.9.1.0 255.255.255.0" on the R6(s0/0)
adds 148.9.1.0/30 but not the 148.9.1.8/30. Actually summary is not allowed.
>
> Any idea ?
>
> Thanks
>
> Kafkaf
>
>
>
> ---------------------------------
> Do you Yahoo!?
> Yahoo! Mail Plus - Powerful. Affordable. Sign up now
> .
---------------------------------
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now
.
This archive was generated by hypermail 2.1.4 : Sat Feb 01 2003 - 07:33:42 GMT-3