RE: Redistribution Rule

From: Huan Pham (Huan.Pham@peopletelecom.com.au)
Date: Tue Oct 14 2008 - 21:37:55 ART


Hi Roger,

I like the way Scott explains. The analogies make it easy to understand.
However, the below rule should not be generalized to all routing
protocols:

"When the ONLY thing you are doing is redistributing RIP. This will look
in the RIP Database and pull routes. If there's no other criteria, then
it gets redistributed to OSPF."

A RIP speaking router only adveritizes all RIP routes to neighbors, so
the RIP database is not much different from the list of RIP routes in
the routing table (the only difference is connected routes)! In this
particular case (redistribution from RIP to OSPF), he is correct.

But if we advertize from OSPF to RIP, we CANNOT pull all routes in the
OSPF database, and advertize them into RIP! The OSPF database could be
very different from the list of OSPF routes in the routing table!
Basically OSPF router advertizes (share) all its database with the
neighbor, including those subnets that are not OSPF routes (e.g. those
installed as EIGRP routes).

Let's take this simple scenario.

Topology as below, R3 redistributes from OSPF to RIP. R2 will see R1
loopback as a RIP route.

    OSPF RIP
R1 ------- R3 ------- R2

If now we run EIGRP on R1 loopback, and between R1 and R3, then R3 sees
R1 loopback as an EIGRP route (no longer an OSPF route).

Even though R3 still have R1 loopback in its OSPF database, it will not
advertize that route to R2 via RIP. R2 will no longer see R1 loopback in
its routing table.

Before introducing EIGRP
------------------------

Serial1/3 23.0.0.3 YES manual up
up
Loopback0 3.3.3.3 YES manual up
up
R3#sh run | s router
router ospf 1
 log-adjacency-changes
 network 13.0.0.3 0.0.0.0 area 0
router rip
 version 2
 redistribute ospf 1 metric 8
 network 23.0.0.0
 no auto-summary

R3#sh ip int brief | ex unass
Interface IP-Address OK? Method Status
Protocol
Serial1/2 13.0.0.3 YES manual up
up
Serial1/3 23.0.0.3 YES manual up
up
Loopback0 3.3.3.3 YES manual up
up
R3#sh run | s router
router ospf 1
 log-adjacency-changes
 network 13.0.0.3 0.0.0.0 area 0
router rip
 version 2
 redistribute ospf 1 metric 8
 network 23.0.0.0
 no auto-summary
R3#sh ip route ospf
     1.0.0.0/24 is subnetted, 1 subnets
O 1.1.1.0 [110/65] via 13.0.0.1, 00:04:42, Serial1/2
R3#sh ip ospf database

            OSPF Router with ID (3.3.3.3) (Process ID 1)

                Router Link States (Area 0)

Link ID ADV Router Age Seq# Checksum Link
count
1.1.1.1 1.1.1.1 295 0x80000004 0x000E50 3
3.3.3.3 3.3.3.3 458 0x80000001 0x003536 2

R2#sh ip int brief | ex unass
Interface IP-Address OK? Method Status
Protocol
Serial1/1 23.0.0.2 YES manual up
up
Loopback0 2.2.2.2 YES manual up
up
R2#sh run | s router
router rip
 version 2
 network 23.0.0.0
 no auto-summary

R2#sh ip route rip
     1.0.0.0/24 is subnetted, 1 subnets
R 1.1.1.0 [120/8] via 23.0.0.3, 00:00:26, Serial1/1
     13.0.0.0/24 is subnetted, 1 subnets
R 13.0.0.0 [120/8] via 23.0.0.3, 00:00:26, Serial1/1

After introducing EIGRP
-----------------------

R1#c
Enter configuration commands, one per line. End with CNTL/Z.
R1(config)#router eigrp 100
R1(config-router)#no auto
R1(config-router)#net 1.1.1.1 0.0.0.0
R1(config-router)#net 13.0.0.1 0.0.0.0
R1(config-router)#

R3#c
Enter configuration commands, one per line. End with CNTL/Z.
R3(config)#router eigrp 100
R3(config-router)#no auto
R3(config-router)#net 13.0.0.3 0.0.0.0
R3(config-router)#

R3#sh ip route | in 1.1.1.0
D 1.1.1.0 [90/2297856] via 13.0.0.1, 00:01:38, Serial1/2

R3#sh ip ospf database

            OSPF Router with ID (3.3.3.3) (Process ID 1)

                Router Link States (Area 0)

Link ID ADV Router Age Seq# Checksum Link
count
1.1.1.1 1.1.1.1 610 0x80000004 0x000E50 3
3.3.3.3 3.3.3.3 773 0x80000001 0x003536 2

R2#sh ip route rip
     13.0.0.0/24 is subnetted, 1 subnets
R 13.0.0.0 [120/8] via 23.0.0.3, 00:00:22, Serial1/1

-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
Roger RPF
Sent: Wednesday, 15 October 2008 8:33 AM
To: 'naman sharma'; 'Cisco certification'
Subject: AW: Redistribution Rule

Hi Naman,

I know how you feel, a had exact the same questions as you, which made
me crazy :o) Scott Morris gave me once a very good explanation on how
the rule works which really helped me to understand it.
I'll just post here the answer which Scott gave me, so all thanks go to
him, it's not my description...
I hope this will also help you

@Scott, I hope I do not violate any rule of "intellectual property
right", posting this message :o))

"When the ONLY thing you are doing is redistributing RIP. This will look
in the RIP Database and pull routes. If there's no other criteria, then
it gets redistributed to OSPF.
 
Now you have introduced extra criteria. So it's not just pulling things
from the RIP Databse, but also asking "Is this really a RIP route"
and if you do "show ip route (x.x.x.x)" it will tell you what it's
learned by. Since this is via Connected, AND you have a "redistribute
connected"
command, that route cannot be added via "redistribute rip", instead it
has to go through your connected route-map. And if it doesn't match, it
won't go.
 
Think of it this way.... If you only own one car, and you want to drive
to the store, there's not much choice that happens in terms of which car
should you use! However, if you own two cars, then you get to choose. Do
you choose by gas mileage? Do you choose based on which car is most
likely to pick up chics? Whatever makes sense, but you have multiple
criteria now.
The router works in the same fashion."

regards

Roger

-----Urspr|ngliche Nachricht-----
Von: nobody@groupstudy.com [mailto:nobody@groupstudy.com] Im Auftrag von
naman sharma
Gesendet: Dienstag, 14. Oktober 2008 23:02
An: Cisco certification
Betreff: Redistribution Rule

Hi Group,

I am just too confused with redistribution default rule, can anyone help
me on this.

As per the rule when we redistribute between 2 protocols then the
connected interfaces on which the protocol is already running will get
redistributed into another protocol immediately.

Now in case one protocol has redistribute connected with route-map
configured this rule breaks, here is the example.

Mutual Redistributing between RIP and OSPF.
RIP running on serial0/0/0 and fa0/0 of router OSPF running on s0/0/1
and Fa0/1 of router.

Now in case there is a redistribute connected command configured under
OSPF with route-map then when RIP is getting redistributed into OSPF
then s0/0/0 and Fa0/0 connected interfaces will not be automatically
redistributed as per the break in the rule and we have to manually add
these interfaces under OSPF redistribute connected route-map.

Please correct me if i am wrong.

Thanks

Blogs and organic groups at http://www.ccie.net



This archive was generated by hypermail 2.1.4 : Sat Nov 01 2008 - 15:35:21 ARST