From: Rick Stephens (rstephens@xxxxxxxxxx)
Date: Wed Jun 13 2001 - 11:01:45 GMT-3
I apologize folks, upon re-thinking my statement below I must ammend my
conclusion and raise a question.
If R1 has lo0 (192.168.1.1) with /30 mask.
If R1 has e0 with (192.168.2.1) /30 mask.
The network for lo0 will be advertised out e0.
But, albeit poor design,
The connected remote router R2 has (192.168.2.2) /27 overlapping mask, R1
and R2 would still communicate.
Since IGRP does not send the mask, R2 will still receive the 192.168.1.0
network.
Does R2 insert 192.168.1.0 /27 in its routing table?
After all, it doesn't know what R1 sent, only that it received a route and
that it SHOULD match its receiving interface.
This is a little tricky for me, I hope I have not completely missed the
issue.
-----Original Message-----
From: Rick Stephens [mailto:rstephens@wantec.com]
Sent: Wednesday, June 13, 2001 8:29 AM
To: 'Mike Gutknecht'; 'GroupStudy'
Cc: 'Carolyn Camarda'
Subject: RE: IP Addressing & IGRP Question
Good explanation, thank you, I did not fully understand.
Then, for all networks to be sent out a single interface, all interfaces
need to have a matching subnet masks. And, for those routes to be accepted
by the remote router it needs to match the subnet mask on its receiving
interface but not necessarily on each of its interfaces.
As for the loopback:
If you had a loopback0 interface with a /30 mask
And an Ethernet0 interface with a /30 mask
And an Ethernet1 interface with a /27 mask
The Loopback would match the E0 mask and be advertised out the E0.
If lo0 is advertised out E0, and the remote router connected interface is a
/30 it will receive the route.
If lo0 is advertised out E0, but the remote router connected interface is
NOT /30 it will ignore the route.
Also, since Loopback0 (/30) does not match E1 (/27) the network will NOT be
advertised out E1.
Is this right?
Example,
int lo0
ip address 192.168.1.1 255.255.255.252
int e0
ip address 192.168.2.1 255.255.255.252
int e1
ip address 192.168.3.1 255.255.255.224
router igrp 100
network 192.168.1.0
network 192.168.2.0
network 192.168.3.0
-----Original Message-----
From: Mike Gutknecht [mailto:mike@justcisco.com]
Sent: Tuesday, June 12, 2001 8:56 PM
To: 'Rick Stephens'; 'GroupStudy'
Cc: 'Carolyn Camarda'
Subject: RE: IP Addressing & IGRP Question
On Rick's last statement,
Maybe also look for a loopback address that is a different subnetmask? They
all need to be the same.
This is incorrect. See Doyle, Page 709. Obviously, this is hideous design
practice, but IGRP or RIP will function.
-Mike G
-----Original Message-----
From: Rick Stephens [mailto:rstephens@wantec.com]
Sent: Tuesday, June 12, 2001 10:32 AM
To: 'GroupStudy'
Cc: 'Carolyn Camarda'
Subject: RE: IP Addressing & IGRP Question
IGRP does not support VLSM, but it does support FLSM. All subnet masks must
be of a fixed length. Since RIPv1 and IGRP routing updates do not include
subnet mask information, a router will assume that the subnet mask with
which it has configured on all its interfaces is the same for all subnets. A
single mask must be used for all subnets of a given classful network.
Different masks can be used for different classful network addresses.
(Caslow p.300)
I think Carolyn said that she tried with a /19 and it didn't work, but tried
it again with a /30 and it did work. The question then would be is does IGRP
support supernets (shorter then the classful boundry). Louie confirmed in
his lab that a /19 can be used.
Confirm 'ip classless'
Maybe also look for a loopback address that is a different subnetmask? They
all need to be the same.
Just my $.02
-----Original Message-----
From: ElleJf(Yahoo) [mailto:ellejf@yahoo.com]
Sent: Tuesday, June 12, 2001 9:43 AM
To: louie kouncar
Cc: 'GroupStudy'; 'Carolyn Camarda'
Subject: RE: IP Addressing & IGRP Question
I would be surprised if /19 mask can work. IGRP does not support VLSM, so
that you can use 192.168.33.0/19 which is a classless subnet. This ip
address belong to class C, so the mask should be /24. In IGRP routing
update packets, only 192.168.33.0 will be recorded (In fact, 192.168.33, 3
octets.) There is no info that you can know what the mask will be.
The only way it would work is that router use locally configured subnet mask
on interface, in this example, /19 as mask. But it doesn't make sense. Can
you provide your config and network diagram?
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com]On Behalf Of louie
kouncar
Sent: Tuesday, June 12, 2001 1:18 PM
To: 'Carolyn Camarda'; 'GroupStudy'
Subject: RE: IP Addressing & IGRP Question
Well,
I just did a test to see if this is an issue, everything worked just
fine using 192.168.33.1/19 on R1 and 192.168.33.2/19 on R2....
Please post your full configs so I can look for issues....
Thanks
Louie Kouncar
UUNET
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
Carolyn Camarda
Sent: Monday, June 11, 2001 11:21 PM
To: GroupStudy
Subject: IP Addressing & IGRP Question
I'm having a mental block with some IGRP configuration, I would
appreciate a
little nudge. I have my router e0 interface set up with the IP address
of
192.168.33.1/19. I have an IGRP network statement of 192.168.33.0.
With
this I do not get IGRP to start up on the E0 interface. I tried the
network
statement of 192.168.32.0 thinking that it may be wanting the VLSM
network
address. None the less I can not get IGRP to fire up.
If I use a class a (10.) or class b (172.16.) address with the /19
mask, I
don't have a problem. When I change the 192 IP address on e0 to a
longer
mask, such as /30 it works fine. My code version is 12.0.5T. I have ip
classless defined on the 2500 router
Is there something with regards to IGRP when the mask can't be shorter
than
the classful boundary to which the address belongs. Or what am I doing
wrong?
Any help would be appreciated.
Carolyn
**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
**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 - 10:31:23 GMT-3