From: Walter Chen (wchen@xxxxxxxxx)
Date: Wed Apr 18 2001 - 11:17:29 GMT-3
Thanks Louie, and Glen. Now I understand how it works and the IOS has no
"bugs". However, I'd still regard advertising a /24 network as a /25
network a "bug" (or at least an undesirable or a not-well-designed feature
I'd say) as the router takes the longer mask of a secondary network and
apply it to the primary network configured on the same interface. Since
IGRP is Cisco's own kid, they should have designed it better even from the
classful routing protocol point of view. But anyway, this is just another
little annoyance I have to remember.
Walter
-----Original Message-----
From: louie kouncar [mailto:lkouncar@UU.NET]
Sent: Wednesday, April 18, 2001 9:27 AM
To: 'Walter Chen'; 'Glen Lavers'; 'Wang, Roger'; ccielab@groupstudy.com
Subject: RE: IGRP route advertisement
Walter,
In the example you provided, the routers are doing exactly what they should,
this is a perfect example of how routing is performed and the IOS has no
bugs whatsoever, In real life you might run into this issue because you are
limited to the IP address range that you have allocated for you, but in a
lab or testing environment, you can see that this will work perfectly.
If this is a real life scenario, I suggest that you use EIGRP.
Good luck...
Louie J. Kouncar (CCIE)/Written
TCO3 Senior Data Center Engineer
UUNET
W-703-343-6645
C-703-304-2460
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com]On Behalf Of
Walter Chen
Sent: Tuesday, April 17, 2001 6:00 PM
To: 'Glen Lavers'; Wang, Roger; ccielab@groupstudy.com
Subject: RE: IGRP route advertisement
Here is a bazzar example on IGRP which made me think that the Cisco IOS may
have some big holes somewhere. I'd welcome anyone who're familiar with IOS
bug report to let me know if they see this problem before.
I used 172.16.x.x address space, and under the interfaces/links I only
specify the last two octets of the IP address such as .12.0/24 (which means
172.16.12.0/24, and the specific address does not matter here). And "sec"
means secondary address configured for the interfaces/links.
lo0 (R3) e0 ----- fa0/0 (R11) s2/1 ------ s2/1 (R12) lo0
.15.0/24 .12.0/24 .11.0/25 .13.0/25
.11.128/25 sec .14.0/24 sec
Every router runs IGRP 1 with one network statement 172.16.0.0 and here are
the routing tables on each router:
On R3:
172.16.0.0/16 is variably subnetted, 4 subnets, 2 masks
C 172.16.11.128/25 is directly connected, Ethernet0
C 172.16.12.0/24 is directly connected, Ethernet0
I 172.16.14.0/25 [100/80225] via 172.16.12.11, 00:00:53, Ethernet0
C 172.16.15.0/24 is directly connected, Loopback2
Note: The only route passed over by IGRP is .14.0/25 which is the secondary
address on link between R11/R12 configured as .14.0/24!
On R11:
172.16.0.0/16 is variably subnetted, 6 subnets, 2 masks
C 172.16.11.128/25 is directly connected, FastEthernet0/0
C 172.16.12.0/24 is directly connected, FastEthernet0/0
I 172.16.13.0/25 [100/80625] via 172.16.11.12, Serial2/1
C 172.16.14.0/24 is directly connected, Serial2/1
I 172.16.15.0/25 [100/610] via 172.16.12.3, FastEthernet0/0
C 172.16.11.0/25 is directly connected, Serial2/1
Note: This router sees every route from both sides except that the .15.0/24
network configured on R3's loopback is passed over as .15.0/25!!
On R12:
172.16.0.0/16 is variably subnetted, 5 subnets, 2 masks
I 172.16.11.128/25 [100/80135] via 172.16.11.11, Serial2/1
C 172.16.13.0/25 is directly connected, Loopback1
C 172.16.14.0/24 is directly connected, Serial2/1
I 172.16.15.0/25 [100/80635] via 172.16.11.11, Serial2/1
C 172.16.11.0/25 is directly connected, Serial2/1
Note: This guy gets the secondary network .11.128/25 configured on the
R3/R11 link correctly, and it also gets the .15.0/25 from R11! The .12.0/24
network did not get here from the .14.0/24 secondary address configured on
the serial lnk to R11.
Anyone has any clue what's going on here??? How can a /24 network be
advertised to the next IGRP router as a /25 network???? I've rebooted all
the routers just to make sure it's not some kind of trasient hicups of the
routers/IOS.
The concerned routers are
R3: 2502 running IOS 11.2(18)
R11: 3640 running IOS 12.1(1a)T1
R12: 3640 running IOS 12.0(4)T
Thanks,
Walter
-----Original Message-----
From: Glen Lavers [mailto:glavers@cisco.com]
Sent: Tuesday, April 17, 2001 3:14 PM
To: Wang, Roger; ccielab@groupstudy.com
Subject: RE: IGRP route advertisement
You can place a secondary address with a /25 on the serial int, it should
then advertise both /24 and /25 addresses. Give it a try, I know I've used
this technique the otherway around for making a poorly addressed VLSM to
FLSM topo work.
Cheers,
Glen
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com]On Behalf Of
Wang, Roger
Sent: Tuesday, April 17, 2001 8:22 PM
To: 'ccielab@groupstudy.com'
Subject: IGRP route advertisement
Hi, all,
I have searched high and low in the archive and can't seem to find anything
about it...
I have a question about IGRP route advertisement. We all know that the
advertisements from IGRP are only with the mask of the interface:
172.16.77.4/25--e0(R4)s0--172.16.66.4/24
R4 is running IGRP. Is there way so that the /25 subnet (e0) can be
advertised out interface s0 (on /24 subnet)? The other end of s0 is a
router running both IGRP and OSPF where redistribution occurs.
Thanks,
Rog
**Please read:http://www.groupstudy.com/list/posting.html
**Please read:http://www.groupstudy.com/list/posting.html
**Please read:http://www.groupstudy.com/list/posting.html
**Please read:http://www.groupstudy.com/list/posting.html
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 10:29:49 GMT-3